Web: http: // www. pearl - hifi . com 86008, 2106 33 Ave.

SW, Calgary, AB; CAN T2T 1Z6

E-mail: custserv @ pearl - hifi . com Ph: + .1.403.244.4434 Fx: + .1.403.245.4456
Inc.
Perkins Electro-Acoustic Research Lab, Inc. ❦ Engineering and Intuition Serving the Soul of Music

Please note that the links in the PEARL logotype above are “live”
and can be used to direct your web browser to our site or to
open an e-mail message window addressed to ourselves.

To view our item listings on eBay, click here.

To see the feedback we have left for our customers, click here.

This document has been prepared as a public service . Any and all trademarks
and logotypes used herein are the property of their owners.

It is our intent to provide this document in accordance with the stipulations with
respect to “fair use” as delineated in Copyrights - Chapter 1: Subject Matter and
Scope of Copyright; Sec. 107. Limitations on exclusive rights: Fair Use.

Public access to copy of this document is provided on the website of Cornell Law School
at http://www4.law.cornell.edu/uscode/17/107.html and is here reproduced below:

Sec. 107. - Limitations on exclusive rights: Fair Use

Notwithstanding the provisions of sections 106 and 106A, the fair use of a copyrighted work, includ-
ing such use by reproduction in copies or phono records or by any other means specified by that section,
for purposes such as criticism, comment, news reporting, teaching (including multiple copies for class-
room use), scholarship, or research, is not an infringement of copyright. In determining whether the use
made of a work in any particular case is a fair use the factors to be considered shall include:

1 - the purpose and character of the use, including whether such use is of a
commercial nature or is for nonprofit educational purposes;

2 - the nature of the copyrighted work;

3 - the amount and substantiality of the portion used in relation to the copy-
righted work as a whole; and

4 - the effect of the use upon the potential market for or value of the copy-
righted work.

The fact that a work is unpublished shall not itself bar a finding of fair use if such finding is made
upon consideration of all the above factors

♦ PDF Cover Page ♦

♦ Verso Filler Page ♦
 Tube Coolers and Equipment Reliability
TUBE COOLER is a device that reduces the
A operating temperature of the enclosing, glass
envelope of an electron tube. While some elements
within the envelope are required to operate at high
temperatures, a surrounding enclosure of conven-
tional, “soft” glass is quite vulnerable to elevated tem-
peratures. Through the action of several mechanisms
extensively outlined herein, this weakness can result
in drastically reduced tube life.
Although common knowledge in other sectors of
the electronics industry for decades, this important
information seems to have escaped the attention of all
designers of tube-type audio gear. We have, with just
one notable exception, never seen tube coolers factory-
fitted to any audio equipment modern or vintage.
The glass envelope, which must act as a high
quality vacuum container, is required to perform
several functions: 1
• it must withstand high operating temperatures,
substantial attendant temperature gradients and
consequent physical stress; without failure.
• at high temperatures, it must resist the pres-
sure differential between the internal high-
vacuum and external atmospheric pressure
• it must be chemically inert, neither adsorbing
gases during manufacture nor liberating them
under high-temperature operation
As the foregoing is essentially a description of an
ideal material, real problems must be anticipated if
soft glass is expected to maintain a high vacuum
over a long period under the conditions imposed by
typical vacuum-tube operation.
While most types of glass are quite stable at low
temperatures, soft glasses become “porous” and
begin to outgas with temperature increases. Where
long working-life is a primary consideration, over-
temperature operation is the main reliability issue
encountered in the operation of soft-glass enclosed
vacuum tubes.
Tube manufacturers, the military and many
large, commercial users have long been aware of the
hazards of such operation and the benefits of reduc-
tions in bulb temperature. Tube makers in particular

Page 1.3 1
are aware that tube life is an inverse function of bulb
temperature.2, 3 (see also Appendix 3)
The MULLARD OSRAM VALVE CO. in England
printed the lettering and the Gold Lion logo on their
famous KT 66 -77-88 series with a temperature-sensi-
†
tive lacquer that would change colour on any tube
run over-temperature. Tubes run too hot were then
permanently indicated to be “…unsuitable for fur-
ther reliable service”.4 (see also Appendix 1)
The basic concept of glass envelope tube cooling
was developed in the 1950s by International Elec-
tronic Research Corp. of Burbank, California work-
ing in conjunction with several branches of the US
military, Cornell University Engineering Laborato-
ries, various tube manufacturers and numerous large
corporations. As a result of extensive research into
causes of equipment failure and the remedies
required, a large number of technical articles
appeared in the literature of the period. Using this
substantial and well documented body of work as a
starting point, PEARL has developed a new and highly
efficient type of cooler for simple, straightforward,
retrofit installation to most existing audio equipment.
AN HISTORICAL & TECHNICAL
O V E RV I E W
With increasing equipment complexity during and
after WW II, the causes of equipment failure came
under intense scrutiny from a number of agencies. The
commercial airlines in the USA formed and main-
tained the non-profit organization, Aeronautical Radio
Inc. (ARINC) to coordinate the development of elec-
tronic equipment for their use, both ground and air-
borne.6, 7 The military in particular, became very dis-
satisfied with the overall rate of equipment failure it
was enduring. The cost of the ongoing maintenance
effort required to keep its vast quantities of equipment
safely and reliably operational grew to enormous pro-
portions. Seeking to alleviate these problems, numer-
ous tube manufacturers were contracted to produce
studies that would detail the reasons for equipment
failure in general and tube failure in particular.8, 9, 10
†Tempilaq is such a product and is available in North America from: TEMPIL,
Div of Big 3 Industries, S. Plainfield, N.J., 201 757 8300


The consensus of this work was that while resistor and 200 Tube lots 5670 / 2C51
capacitor failures accounted for approximately 7% of 100
1
failures, an amazing 75% of failures were due to tubes. 2
80
Subsequent, detailed investigations carried out by

Survival Percentage
numerous, widely separated researchers revealed that 60 Curve 1 - 5.04V
tubes will fail in a radically premature manner when Curve 2 - 6.30V
3 Curve 3 - 7.56V
forced or simply allowed to operate at excessive bulb 40
temperatures. 10 (see also Appendix 3: ref ’s 7, 8, 9, 10, 11, 12)
A study of over 150,000 tubes of 20 different 20

types undertaken by ARINC lists a number of proce-

dures that increase the reliability of vacuum tubes.8 0
0 1000 2000 3000 4000 5000
Foremost amoung these measures is the operation Operating time in hours
of tubes in a manner that reduces bulb temperature.
In order of decreasing adverse effect on tube life, 200 Tube lots 5670 / 2C51
excessive bulb temperature causes: 7000

Transconductance in µmhos
Curve 1 - 5.04V
6000
• the evolution of gas within the tube, which Curve 2 - 6.30V
causes the steady reduction of transconduc- 3 Curve 3 - 7.56V
5000 1
tance. Left unremedied, this process can cause 2

the tube to glow with a lovely electric blue 4000

colour while acting as a forward biased diode.
• in part, the development of an interface resis-
tance between the surface of the nickel tube
that forms the body of the cathode and its
electron emitting oxide-coating. This effect is
partially the result of over-temperature opera-
tion of the cathode and can be caused by:
- excessive filament current,
- excessive overall operating temperature,
- in some types, long periods of operation in a
cut-off condition resulting in the develop-
ment of such a high value of resistance that
current flow will not restart when the tube is
biased so as to resume current flow. This
was a problem with the famous ENIAC (Elec-
tronic Numerical Integrator and Computor)
developed as part of the Manhattan Project
during WW II. Special tube types, 6SN7GTB
& 5692 for example, were developed for such
applications. Interface resistance is also
responsible for reductions of transconduc-
tance. 10, 11, 12, 13 See Fig. 1.
• grid emission, a prime factor in the noise
increases seen as tubes age. A slow accumula-
tion of cathode material on the grid wires ini-
tiates an ever increasing, low-density electron
flow from the grid to the plate. Flowing from
ground through the grid resistor, this fluctuat-
ing current develops a noise voltage that
appears between the grid and signal-ground.
Being thereby applied to the input of the tube
in the usual way, this noise voltage is likewise
amplified in the usual way.
By parts, the deposition process is an outcome
of the water cycle, later described, operating in
tubes run over-temperature and also the result
2
3
3000
2000
0 1000 2000 3000 4000 5000
Operating time in hours
Fig. 1. Tube transconductance and working-life are shown plotted
as a function of filament voltage. Note that operation of this particu-
lar tube at its rated filament voltage results in a 27% drop in
transconductance during the first 1500 hrs. of operation. Operation
of the filament at a 20% reduction in voltage results in very nearly
constant transconductance over a much extended working-life.
of full B+, cold-cathode startups.
• cathode poisoning, resulting in a premature
reduction of electron emitting capacity (per-
veance). 14, 15 The authors of these references
state that:
“A new and unexpected source of cathode poi-
soning gas is seen to derive either directly or
indirectly from the heated glass envelope. Such
gas is more destructive in action than any of the
normal gas so far examined This gas is believed
to be water vapour which has been shown to
have dire effects on cathodes operating in the
vicinity of 725 °C.”
• migration of both the getter patch and
unflashed getter material, another likely result
of the water cycle.
• interelectrode leakage, whereby voltages
impressed upon specific elements within the
tube wrongly appear on other electrodes. This
can be caused by water cycle induced migra-
tion of conductive getter metals onto the insu-
lating micas and to the base of the envelope,
where the pins exit the tube, causing lowered
resistance among the tube’s elements.
• contamination, resulting in tiny bits of mater-

Page 1.3 2

 ial coming adrift within the envelope.
• glass failure, with attendant loss of vacuum if
not outright failure.
• increased grid temperature that can result in
increases in the normal thermionic emission
from control grids. In poorly designed circuits
this effect can cause grid runaway, where nor-
mal grid bias is lost and plate current rises to
saturation levels in an uncontrolled manner.
The typical outcome is the rapid demise of the
tube in question.
The glass envelope of a tube presents a
significant resistance to the flow of heat from the
tube.16 Glass is a very poor thermal conductor and is
virtually opaque to thermal radiation at tempera-
tures below 400 ° C. Consequently, the envelope
absorbs nearly all of the heat radiated from the ele-
ments contained therein with the result that a hot-
spot occurs in the glass adjacent the centre of the
plate, the hottest part of the the tube structure. This
hot-spot causes a substantial temperature gradient
along the length of the envelope that can result in the
centre of the envelope running 25 to 100°C. hotter
than either of the cooler ends. This creates enormous
physical stress within the glass and in extreme cases
can result in failure. See Fig. 2.
An interesting piece of work done by Rogers
Majestic Co. of Toronto in 1933 suggests a mecha-
nism whereby something like atomic osmosis occurs
vapour, carbon dioxide and nitrogen. With further
increases in temperature this process dramatically
intensifies and usable life is sharply reduced by the
initiation of a water cycle described below.18
Water vapour is dissociated by the hot internal
elements within the lamp or tube into hydrogen and
oxygen. As well as reacting with the cathode material
of the tube, the oxygen will react with the other
metallic surfaces such as the getter patch†† and any
unflashed getter material, with some of these areas
being hot enough to blow off an oxide that can then
deposit on various cooler surfaces. Meanwhile, the
highly reductive hydrogen migrates to the deposit,
reduces it—liberates the oxygen from the oxide—
and leaves the deposit as recombined water vapour
ready to begin the cycle all over again.
In its ionized form, the hydrogen acts to reduce
the electrical resistivity of the vacuum by turning it
into a partially conductive medium, thereby subtly,
yet continuously, decreasing the tube’s ability to
accurately control electron flow through itself.
While gas penetration and out-gassing are
significant mechanisms by which the vacuum is
spoiled, excessive temperatures can cause various
gases to evolve from the elements within the tube
2.00
3.71 WATTS DISSIPATION
4.56 SQUARE INCHES 5751 / 12AX7
0.814 WATTS PER SQUARE INCH
1.75
27°C AVERAGE AIR TEMPERATURE
- - - 90° ROTATION SURVEY
1.50 •1 •7
Position (Inches)

at high glass temperatures.17 The hypothesis is that 1.25

•2 •8
the sodium in the glass becomes mobile and acts as 1.00
an electrolyte thereby facilitating the bodily migra- •3 •9
0.75
tion of atoms of atmospheric oxygen through the •4 • 10
glass and into the tube. These reduce the usable life 0.50
•5 • 11
of the cathode by combining with its electron-emit- 0.25
•6 • 12
ting surface, creating oxides that reduce its effec- 0
tiveness. It is likely that other gases evolve from this 60 80 100 120 140 160
poisoning action and these can adversely effect the Bulb temperature, degrees Celsius
hard vacuum upon which the tube relies for linear
and effective operation.† 2.00
12.57 WATTS DISSIPATION
5687 WA
4.56 SQUARE INCHES
Noted by workers in the incandescent lamp 1.75 2.75 WATTS PER SQUARE INCH
39°C AVERAGE AIR TEMPERATURE
Similar to
industry is the fact that glass envelopes will begin to - - - 90° ROTATION SURVEY 6CG7 / 6FQ7
1.50 •1 •7
Position (Inches)

outgas into the evacuated volume when the surface

1.25
temperature exceeds 100°C. The evolution of gas is •2 •8
1.00
bought about by chemical decomposition and release •3 •9
of adsorbed molecules which are, principally, water 0.75
•4 • 10
0.50
•5 • 11
†The need for thorough bulb evacuation escaped the originator of the triode, 0.25
Lee deForest. It was not until his successors started pulling a genuinely hard •6 • 12
vacuum within the envelope that the wonderfully linear triodes presently 0
enjoyed could be developed. 120 140 160 180 200 220
††The getter patch is the shiny, mirror-like spot seen on the inside of most

receiving tubes. It results from the combustion—for want of a better term—of the
getter, a sacrificial element placed within the tube. After the envelope has been
pumped out and sealed, the getter is heated by RF induction to a high tempera-
ture. The special formulation of the getter combines with any remaining gases
within the tube to form metallic compounds that subsequently boil themselves
onto the inside of the glass wall creating a reflective spot on the glass.
Bulb temperature, degrees Celsius
Fig. 2. Bulb temperature plotted as a function of position on the
bulb. It is interesting to note that the 5751/12AX7, a relatively cool-
running tube, is one of the most reliable and longest-lived of any
used in audio.

Page
 structure itself, particularly from the plate. We sus-
pect that these actions are responsible for the slight
dimunition in sound quality often heard from tubes
run too hot during the first few hundred hours of life
even though conventional “usable life testing” will
indicate that the tube is fine. It must be stressed that
the various mechanisms of deterioration can oper-
ate as a higher-order exponential function of tem-
perature increase. If the tube hot spot temperature
is reduced 25 to 150 ° C . the rate of these various
contaminating and poisoning actions can be
reduced 2 to 50 fold. In other words, small decreases
in bulb temperature often result in seemingly dispro-
portionately large increases in tube life. See Fig. 3.
In dozens of studies, tube life has been shown to
dramatically increase as overall bulb temperature is
reduced. However this must be achieved by means
that substantially eliminate the temperature gradi-
ent along the bulb length. For this reason simple fan
cooling, while effectively reducing the ambient tem-
perature within equipments (and thus the average
temperature of the envelope), cannot approach the
results obtained by a properly designed tube cooler.
200 Tube lots 5654 / 6AK5W
100
125°C
Survival percentage
Fig. 4 illustrates the sort of gradient equalization
required to substantially increase tube life, while
Fig. 5 illustrates the limited value of forced air cool-
ing of bare bulbs. Fig. 6, derived from IERC’S work,
shows that a tube with a properly designed cooler
can run at a lower temperature in a simple convec-
tive environment than does the same tube running
bare in a 500´/min forced air stream. Figs. 1, 2 & 3
were taken from a landmark study of the factors
effecting tube life conducted by W.S. Bowie at GE’s
Owensboro, Kentucky facility in 1954.
We spent 18 months researching the effects of
glass temperature reduction and the development of
our present coolers resulted from that work. We
consider this project to have been successful in that
the sorts temperature reductions and gradient elim-
ination repeatedly shown to substantially increase
tube life have been achieved.
Numerous military, institutional and commercial
studies involving thousands of tubes in scores of oper-
ating environments and conditions have shown that
bulb temperature reductions on the order of those
achieved by our coolers yield tube life in the many thou-
sands of hours. While many of these studies showed
increases in tube life of 5 to 50 times, none showed
improvements less than a doubling of previous life.
Illustrative of the sort of reliability gains obtain-

80
able by reducing bulb temperature is a study under-
60 taken ARINC. Appendix 1, ref 15 This was a two year long
100°C
field observation of an equipment using 6-6005 minia-
192°C
40 ture tubes. Prior to the commencement of the study,
312°C the average-tube-life was under 1000 hours. By simply
20
253°C outfitting the gear with heat-dissipating tube shields
0
the average-tube-life skyrocketed to 12,000 hours.
0 1000 2000 3000 4000 5000 Study of Fig. 3 will reveal the plausibility of this result.
Operating time in hours Several manufacturers of audio equipment list the
expected life of power amplifier output tubes in the 1-
200 Tube lots 6005 / 6AQ5W 2000 hr. range. By the simple application of tube coolers
100
these short life expectancies can be at least doubled.
220°C
Being a low-cost, reusable, one-time investment, tube
80
coolers can yield substantial savings, not only in tube
Survival percentage

60
237°C
replacement costs but by way of a reduction in concern

40
Low Bulb Temperature High
20

316°C 261°C
347°C Bare bulb temperature
0
0 1000 2000 3000 4000 5000
Operating time in hours
Bulb temp with cooler
Fig. 3. Taken from GE’s major study of vacuum-tube life done in
1954, these data illustrate the extremely harmful effects of elevated
temperatures It is interesting to note just how very abruptly tubes will
fail when operated at unreasonably elevated temperatures. For
instance, after 2500 hours of operation, a mere 15% of a batch of
6005 tubes run at 261° C. are operational while 90% of those run 41
degrees cooler at 220° C. survive 5000 hours. That is the equivalent
of 6 times as many tubes living twice as long. Fig. 4 . Tube-surface temperature gradient with proper cooling.

Page 1.3 4

over the condition of the tubes in ones equipment. With
tube gear costing what it does, no one’s needs are well
served by ongoing tube deterioration and the need for
costly tube replacement every year or two
A F E W P A RT I C U L A R S O F D E S I G N
It is somewhat unusual for a manufacturer to
mention the products of other suppliers. However
we feel a definite obligation to provide information
about IERC’s coolers as they were producing and
selling their devices, by the hundred thousand, well
before we took up our careers in audio.
The differences between our product and IERC’s
are significant and relate largely to the fact that
printed circuit boards are a way of life in electronic
manufacture today whereas during the 1950s, point-
to-point wiring within a metal chassis was the order
of the day.
As IERC’s coolers are chassis-mounting devices,
employing the ample sink provided by the equip-
ment’s metal work to dissipate heat, they are not well
suited to installation on pc-board constructions.
PEARL coolers dissipate heat directly into the
atmosphere from a large area, radial fin arrange-
ment and are not dependent upon the presence of
any additional heatsinking metalwork. See Fig. 7.
2 1
50 75 100 125 150 175 200 225 250
122 167 212 257 302 347 392 437 482
Fig. 5. Bulb temperature-gradient is shown for an EL34 running at
maximum dissipation. Curve 1 is for a tube running in free air with no
forced-air cooling. Curve 2 is for the same tube running in a
100ft/min air stream guided by a plastic chimney, as shown.
Note that while the average temperature of the bulb is reduced, the
substantial gradient is essentially unchanged. The use of a chimney
to direct a low-speed air-flow over the tube is a much better
approach than gently blowing air, from some distance, at a tube. The
temperature reduction, as shown here, is much greater than is oth-
erwise the case. Tubes placed directly in high-velocity air-streams do
show significant average temperature reductions with correspond-
ing life extension.

Page 1.3 5
While this has advantages for modern, station-
ary equipment such as home audio gear, we have
been required to sacrifice the small size and tube
retention capability of the IERC devices. Due to the
conditions of extreme, acceleration, shock and vibra-
tion encountered in early guided missiles and other
military hardware, IERC’s coolers were designed to
hold both the tube and themselves firmly in position.
As we don’t expect that anyone who owns high per-
formance audio gear is going to hammer it pell mell
down 40 miles of washboard road in the back of a 4 x
4, we have designed and optimized our coolers to do
three things; retrofit easily, effectively dissipate heat via
convection and radiation and help damp the internal
resonances that define the degree of microphonic out-
put a tube will exhibit. Because of our differing design
goals we have been able to concentrate quite single-
mindedly on maximizing heat removal from the bulb.
In order to effectively sink heat from the glass
envelope into any cooling device, a significant
amount of metal must be placed in contact/close
proximity with the glass surface.19 This is rather
difficult to do because the glass envelope of a tube is
neither perfectly round nor parallel-sided.
Further complicating this situation is the need for
firm metal-to-glass contact in spite of significant vari-
Clear acrylic
plastic chimney
275° C
527° F Chassis
Air flow


 12BY7 Thermal equivalent dissipating 10 watts.
200
Bright
180
JAN s
hield
Bulb Temperature (°C)
Black
160 JAN s
hield
140
Black
JAN s
hield
with w
120
Bare b
100
80
IERC T6-1025H
60
0 50 100 150 200 250 300 350 400
Air Velocity (FT./MIN.)
Fig. 6. The operation of a thermal-equivalent tube developed by Cornell Aero-
nautical Labs is shown under various conditions. Note that a bare bulb in a 500
ft/min airstream runs hotter than when properly cooled in still air. The Joint-Army-
Navy—JAN—shield is the (usually) shiny slip-on affair most people recognize as
a tube shield. Note the overall increase in temperature its application always
causes. Such an increase has been repeatedly shown to shorten tube life and for
that reason its use was virtually banned by the US military by 1960. This fact
seems to have escaped the attention of designers & builders of Golden Age
audio gear as it is by no means uncommon to encounter such equipment
employing the worst-case-bright-JAN-shield.
ability in bulb size among different brands of the same
tube type. For instance, the EL34 is found in bulbs of
three different sizes, the 6550 in two sizes and so on.
Power tubes are typically enclosed within 11⁄8″,
1 ⁄4 , 11⁄2″, 13⁄4″, 2″ and 21⁄2″ diameter bulbs, small signal
1 ″
octal tubes within 11⁄8″ & 11⁄4″ dia. bulbs, 9 pin minia-
tures within .80″ to .84″ bulbs and 7 pin miniatures
within .70″ to .73″ dia. bulbs.
The fairly small deviation in size among the 7-
pin and 9-pin miniatures, allows us produce only
one size of cooler to fit each type while achieving the
required metal-to-glass intimacy. The most work-
able solution we have found to the problem pre-
sented by the wide variation in bulb size seen with
power tubes is to produce coolers to individually fit
the various bulb diameters encountered. As we need
to know the diameter of the tube to be cooled, we
supply a special order form that has a direct-read-
ing, diameter-ruler printed at the bottom that is
intended to be snipped off and wrapped around the
tube. The diameter indicated is simply entered on
the form so that we can supply the correct cooler.
PEARL coolers are held against the bulb by a
compressive band arrangement, with round, high-
temperature, elastomeric bands providing the com-
pressive force.
In cases where tubes are placed too close together
to allow the use of the compressive bands they can

Page 1.3 6
often be dispensed with and the coolers will
hold themselves with adequate force against
the glass. This way coolers can be fitted onto
adjacent tubes that are as little as .55″ bulb-
to-bulb from each other.
Our coolers are formed from solid cop-
per foil into a radial fin arrangement that
acts as a static fan to pull cooling air through
indow the fins of the device. Being made from fairly
s
thin material the cooler flexes easily and
thereby accommodates deviation from bulb
ulb roundness and side-wall parallelism.
The coolers are easily fitted by simply
placing the device on its end on a flat sur-
cooler face, pushing the tube into the centre hole
and then sliding the cooler a further 1⁄4″ to 1⁄2″
450 500 onto the bulb as required. The compressive
bands can then be adjusted on the cooler to
give a tidy appearance. In the event that a
large bulb is forced into a cooler of an inap-
propriately small size, the fins will simply
spread open and deform, no damage will
occur to the tube and hence there is no pos-
sibility of receiving nasty cuts from the
ragged edge of a bulb broken by excessive
insertion force.
As a potential purchaser of PEARL tube
coolers you can reasonably expect to get at
least double the life from any new tube no matter how
little power it is dissipating (ie. a 12AX7) and very prob-
ably three to five times the life from hot running power
tubes in output stages where the amplifier is being dri-
ven hard and/or biased to run well into Class A opera-
tion.
Coolers are available to fit straight-sided
envelopes only, with models to fit any such device
from small 7-pin miniatures to fan-cooled applica-
tions involving large transmitting tubes such as the
813, a 100 watt pentode.
SOME POSSIBILITIES FOR SONIC
IMPROVEMENT
Designers may select higher operating currents
in pre-amp stages thereby realizing gains in
transconductance and consequent noise reduction
previously attainable only through ongoing aggrava-
tion and cost to the end-user due to constant tube
deterioration and replacement. Such increases in
current flow also result in a lowering of the plate
resistance, rp , and as a result of the smaller values of
cathode resistor used to bias for increased current
flow, the effective plate resistance, rp′. This lowering
of the voltage amplifier’s output impedance allows
improvements, for instance, in the ability of no-feed-
back line stages to drive the highly capacitive cables
so (unfortunately) prevalent in high-end audio today.


 Fig. 7. The various styles of PEARL coolers are illustrated above.
The types for 7 & 9 pin miniatures and small-signal octals (6SN7,
etc.) are essentially similar, differing only in the number of fins and
the size of elastomeric band used to compress the cooler. The
Power stages may be biased somewhat further into
Class A operation with an attendant sweetening of har-
monic structure and gains in soundstaging, trans-
parency and relaxed musicality These improvements
are particularly impressive in those output stages oper-
ated fixed-screen-voltage pentode/beam tetrode; see
our Audio Notes 2.1 & 2.1.1, A Little Input on Audio
Output Transformers & Update for more information.
Several users have reported being rather per-
plexed by the improvements in sound quality that
manifest when coolers are fitted “on the fly” to equip-
ment that is operating. There is the immediate relax-
ation of the sound quality that we associate with
reduced microphony and there is a further improve-
ment with time as the bulb temperature reduces.
Radiant thermal energy from the plate structure
is largely responsible for heating the glass, with the
glass being a poor thermal conductor. It is reason-
able to think that a build-up of heat in the glass
causes a buildup of heat in the plate. If a cooler is
applied to the glass, resulting in a reduction of glass
temperature, it is reasonable to assume that the
plate temperature will reduce as well.
We have reason to believe that there is a relation-
ship between plate temperature and the thickness of

Page 1.3 7
standard glass-to-glass spacing is .75” or 19mm however when
tubes are too close together the O-rings can sometimes be omitted
allowing the colers to be fitted where clearances are as marginal
as .55˝ or 14mm glass-to-glass.
an electron gas or film that must exist near the sur-
face of the plate. Being in a state of chaotic motion,
the electrons that make up this film will have the
effect of modulating the density of the incoming elec-
tron beam in an unpredictable way. In other words,
this film adds a noise component to the signal being
amplified and does so to a degree that can be shown
to be a function of plate temperature.
The presence of such a film is a near certainty in
that the conditions necessary for its creation, ie.
pure thermionic emission, cathode-splatter-aided
thermionic emission and secondary emission, are
all met by activity at the plate surface. Given the
foregoing, it can be surmised that it’s possible for a
cooler plate structure and a consequently reduced
noise-film thickness to produce less of what could be
called a signal induced noise. There are a great num-
ber of such SIN effects operating in audio devices of
all kinds and their reduction nearly always results in
a sound that is smoother, more natural and less
grainy. Sonic improvements of exactly this sort are
being reported by many users of tube coolers.
There are basically three mechanisms by which
which a stray-electron film could be produced and
these are outlined below:


• Thermionic Emission – If a material is heated in
a vacuum, the presence of an electron cloud
can usually be detected close to the material’s
surface. Thermionic emission has been shown
to increase in intensity as a higher order func-
tion of increasing absolute temperature—mea-
sured in degrees Kelvin, which start with the O°
point being absolute zero.20, 21
In other words, many materials will emit elec-
trons when heated and do so with an intensity
that increases dramatically as temperature is
increased. The outer-orbital electrons of the
atoms of metals are not firmly bound to their
nuclei and are believed to pass from atom to
atom in bucket-brigade fashion, with this prop-
erty widely held accountable for the electrical
conductivity of metals. These loosely bound
electrons are sometimes likened to the mole-
cules of a heated liquid as they are in constant,
unpredictable motion and often leap from the
surface in an attempt to escape from it. Such
departure induces a positive-electrostatically
charged area within the host material that
attracts the negatively charged electrons back to
the surface from which they escaped. The force
of re-attraction set up by the escaping electrons
is called an image force
In order for electrons to make good any
attempt to leave the surface, some energy
must be added to the electrons while they are
contained within the metal—energy that pro-
vides the velocity required for escape and that
will be lost upon the electron’s emergence
from the metallic surface.
The kinetic energy lost in this manner differs
in amount from one material to the next with
the term electron affinity being used to describe
the amount of energy that must be added to the
material, usually by heating, to provide ade-
quate electron escape velocity to overcome the
image force. Materials with low electron
affinity will emit substantial numbers of elec-
trons with only moderate heating while those
exhibiting high electron affinity require much
higher temperatures for equivalent emission.
• Cathode-Splatter-Aided Thermionic Emission –
During the process of tube manufacture some
of the low-electron-affinity coating applied to
the cathode almost invariably makes its way to
the plate and other elements within the tube.22
While the grids can sometimes be run hot
enough to blow-off such deposits, some mate-
rial tends to remain bound to the plate.21
Additional cathode material is normally
stripped off the cathode in circuitry where the
B+ voltages are applied to tubes before the

Page 1.3 8
filament has heated the cathode to a suitable
operating temperature. In this condition of no
appreciable current flow through the tube,
there is no significant voltage drop across the
plate load resistor resulting in the presence of
nearly the full B+ voltage on the plate of the
tube. This voltage typically being at least twice
the normal operating plate voltage, the electro-
static force of attraction from the plate acting
on the cathode coating is likewise at least dou-
ble that seen in normal operation. By these
mechanisms the plate is usually contaminated
with some amount of highly emissive, low-
electron-affinity cathode-coating material.
• Secondary Emission – “When a solid body is sub-
jected to bombardment by electrically charged par-
ticles, some electrons that may be detected under
suitable conditions are always emitted. Although
this process, commonly designated secondary
emission has been seen to occur in various forms,
by far the most widely investigated type is that in
which an electron beam falling upon the surface of
a target in a vacuum causes the emission of a
stream of electrons from the surface upon which it
impinges.” 23.
Secondary emission of electrons from the
plate of a tube results primarily from collisions
between incoming electrons from the
cathode—which may penetrate some distance
into the material—with electrons in the mater-
ial itself. The usual result is that several elec-
trons are kicked out of their orbits within the
plate material and into the space in front of the
plate by every electron striking the surface. The
secondary emission phenomena is unique in
several ways as regards other types of electron
emission activities. It does not appear to be
effected by temperature nor overly much by the
nature of the material being bombarded, with
materials of widely varying characteristics
showing variations in emission of only one
order of magnitude all other things remaining
equal. Conductors and insulators alike show
quite similar patterns of secondary emission.
There are however, a couple of factors effecting
secondary emission that are pertinent to this
discussion. Incoming electron velocity, whether
generated by an accelerating gun or screen-grid
arrangement (as in a pentode) or a positive
potential on the target surface (as on the plate
of a triode) shows a marked effect upon the lib-
eration of secondaries. High incoming-electron
velocity yields high secondary emission.
The angle of incidence at which an incom-
ing electron strikes the target shows a
significant effect, with grazing angles of inci-


 dence yielding two to three times the sec-
ondary emission of normal, 90 ° impact.
In summary:
• the intensity with which thermionic emission
takes place is very much a function of the tem-
perature of the surface and the electron
affinity of the material at the surface. Coatings
as thin as one molecule in depth can have a
radical effect on the operating electron-affinity
of a material.†
• the intensity with which secondary emission
takes place, as it concerns us here, is a func-
tion of incoming-electron velocity and the
angle of incidence at which incoming elec-
trons strike the surface.
Using this information it is possible to form a
reasonable theory that can explain several seem-
ingly peculiar phenomena that effect the sonic sig-
nature of different tube types. The basic idea is that
the presence of a film of electrons at, and likely just
penetrating, the plate surface will have a significant
effect on the sound of a given tube.
This notion may be borne out by a couple of
observations repeatedly made by audiophiles over
the last thirty years.
The so-called ‘flat-plate’ Telefunken 12AX7 has
long been regarded as the best sounding 12AX7 ever
produced. As its name implies, the plates in the Tele-
funken part are flat while other brands have stiffen-
ing ribs pressed into the plates by way of depres-
sions running across each plate. As previously
explained, secondary emission is much greater
when the incoming electron stream strikes obliquely
against a surface.
Incoming electrons striking the flat plate of the
Telefunken style device will produce less secondary
emission than when in collision with the ribbed plate
of the conventional construction, thereby creating a
lower electron-film density in the flat-plate part than in
the ribbed-plate device. Because the 12AX7 is a cool
running, low transconductance tube it likely does not
suffer greatly from thermionic plate emission from
either cause so most of the noise-producing electrons
will be generated by secondary emission.
The second seemingly odd observation is that
tubes with plate structures that are a shiny black
appear to give a cleaner, more open sound than tubes
of the same type with plates that have not been
treated to produce that sort of surface. I presently
believe that such a surface results from graphite
dusting and subsequent hydrogen firing of the plate
†This outcome has been exploited for decades in the creation of low elec-
tron-affinity surfaces for use as cathodes in vacuum tubes.

Page 1.3 9
structure but more research is needed to confirm this
speculation. Because carbon has a high electron
affinity, thermionic emission from the plate will be
lower than for untreated parts. Hence the electron
film producing effect of plate emission will be lower
for a carbon treated plate.
This thought is consistent with the foregoing
ideas in that it strongly suggests that a reduction in
plate emission—either thermionic or secondary—
contributes to an improvement in sound quality.
Although changes in design of tube internals
must be implemented by the manufacturer, there
are a couple of ways that an end-user or designer
can use this thinking to improve the sound of a
given circuit.
The first, and easier of the two, is to reduce plate
temperature by simply fitting a tube cooler.
The second is to carefully reduce the plate volt-
age. Such voltage reductions must be undertaken
with the idea firmly in mind no panacea is at work
and that the method is viable only in low-level cir-
cuitry where input voltages are low and individual
triode gains are moderate—as in the bottom section
of a conventional cascode in a phono input stage.
Attempts to run higher-level stages at low plate
voltages can result in steady-state harmonic and IM
distortion increases that will probably outweigh any
gains this method might offer.
An as-yet-unexplored possibility for improve-
ment is that which may accrue from negative
potential operation of the coolers through 10MΩ
resistors. It may be that the close proximity of a
large area of metal, at some moderate minus—50-
75V—potential, to the envelope will have the effect
of reducing or eliminating regions of static charge
in the bulb although any beneficial effects from this
would likely take several hours, if not days, to man-
ifest because of the loose electrical coupling
between the cooler and the glass.16 Such statically
charged areas result from bombardment by stray
electrons emitted from the cathode. Secondary
emission is known to result from such collisions,
with several electrons being knocked out of the
glass by every impinging stray electron. These spu-
riously emitted electrons ultimately find their way
to the plate where they add a noise component to
the flow of current in the plate circuit.
In the region where secondaries have been dri-
ven out, a region of positive charge is created within
the glass that naturally attracts further cathode-
strays, thus establishing a self-stoking vicious cycle.
In power tubes these areas can become extremely
hot and such temperature rises can cause outright
glass failure. In small signal tubes, these regions tend
to migrate in a somewhat unpredictable fashion and
some odd results can be imagined from this action.


 A F E W P A RT I C U L A R S R E G A R D I N G
THE USE OF PEARL COOLERS
In the great majority situations the installation
of standard PEARL coolers is very simple, requiring
only that the cooler be placed on a flat surface and
the tube forced into it. In cases where tubes are
placed too close together for the standard coolers to
fit without the elastomeric bands fouling each other,
coolers from the close-fitting series can usually be
fitted. Sometimes tubes are so tightly packed that
the radial fins of adjacent coolers must be meshed
like gear teeth. This generally requires a little
fiddling but the effort is always worth it as tubes that
are too close together radiate heat onto each other,
with some very high bulb temperatures and short
tube lives being the outcome.
In some situations it may beneficial to ground the
coolers by soldering a light gauge, flexible lead to the
device and tying that lead to signal ground. Where RF
interference is a problem 1KΩ resistors should be
fitted to both ends of the lead ie. a resistor is soldered
to the cooler with the flex-lead being soldered to the
other end of the first resistor with the second resistor
being soldered to the other end of the flexlead.
Simply fitting coolers into tuners is generally not
a good idea because the added capacitance from the
plate-to-cooler-to-ground or other nearby compo-
nents can upset tuner alignment. While this can usu-
ally be re-tweaked with little difficulty, coolers should
be grounded via short leads before this is attempted.
In pre-amps where one dual triode is handling
both left and right channel signals, the added capaci-
tance between the adjacent plates within the tube can
REFERENCES
1 Tube Manufacturing at Western Electric: The WE300B, A.R.
Balaton, Journal of the Audio Engineering Society, November,
1989
2. High Power Audio Amplifiers, M. Horowitz, Audio, March,
1958.
3 Discussion with the people at MPD Inc., Owensboro, Ken-
tucky. This facility was originally a General Electric plant and
is the same factory mentioned below, 1990.
4 Discussion with the people at English Electric Valve Co., Toronto,
Canada, Canadian distributor for The MO Valve Co., 1989.
5 Heat Dissipating Tube Shields and Their Relation to Tube Life
and Equipment Reliability, J.C. McAdams, IERC, Burbank,
California, 1959
6 High Reliability Miniature Tubes, G. Gage, General Electric
Co., Owensboro, Kentucky, Electronics, December, 1950.
7 Long-Life Tubes for Industry, E.K. Morse, ARINC, Electronics,
September, 1950
8 Military Reliability of Electron Tubes, E.R. Jervis, J. &
Swauger, ARINC., Electronics, April, 1954.
9 Reliability in Miniature and Subminiature Tubes, P.T. Weeks,
Proc. I.R.E., May, 1951
10 Study of Electron Tube Life, W.S. Bowie, General Electric Co.,
Owensboro, Kentucky, 1954, Mil. contract #DA-36 039-SC-42524.
11 Change of Mutual Conductance with Frequency, W. Raudorf,
Wireless Engineering, October, 1949.

Page 1.3 10
cause changes in sound-staging and it is recommended
that coolers be grounded in these cases as well.
There are several situations where coolers may
effect sound quality in a slight way, usually by
changing the shape of the electrostatic fields that
exist within all equipment, tube or solid-state. As
these E-fields are not generally optimized in most
equipment, changes can have unpredictable results.
Sometimes the sound will improve and sometimes
not. If there is any doubt, the best solution is to
ground the coolers.
T H E A LT E R N AT I V E T O P E A R L
COOLERS
As a courtesy to our forebears we publish IERC’s
particulars. We recommend their coolers when the
following conditions prevail:
• the equipment has its tube sockets mounted
directly to or directly below a metal chassis
• the tubes are very close together
• the equipment is subjected to extreme vibra-
tion or shock.
International Electronic Research Corp.,
135 West Magnolia Boulevard,
Burbank, Calif., 915027
Cathy Sparks: Inside Sales Manager
12 Valve Cathode Life, C.C. Eaglesfield, Wireless World, December,
1951.
13 Unusual Tube Effects Cause Circuit Troubles, W.E. Babcock,
RCA, Electronics, Engineering Issue, September 12, 1958
14 The Life of Oxide Coated Cathodes in Modern Receiving
Valves, G.H. Metson, S. Wagener, M.F. Holmes & M.R. Child,
Read before the I.E.E., early 1950’s.
15 Oxide Cathode Life, R.W.H., Publication and date unknown,
copy available from PEARL.
16 Thermal Analysis and Control of Electronic Equipment, A.D.
Kraus, & A. BarCohen, McGraw-Hill, 1983.
17 The Spray Shield Tube, H. W. Parker, & F.J. Fox, Proc I.R.E.,
May 1933.
18 Technical Reference Manual; and subsequent correspondence,
Lamptronix Co., Crystal Lake, Illinois, USA 60014, 1989.
19 Water and Forced Air Cooling of Vacuum Tubes, I.E.
Mouromtseff, Proc. I.R.E., April 1 942.
20 Elements of Thermionics, W.E. Danforth, Proc. I.R.E., May
1951.
21 Studies on Grid Emission, Ga Espersen & J.W. Rogers, I.R.E.
Transactions-Electron Devices, April 1956.
22 Secondary-Emitting Surfaces in the Presence of Oxide-Coated
Cathodes, S. Nevin & H. Salinger, Proc. I.R.E., February, 1951.
23 Fundamentals of Secondary Electron Emission, M.A. Pomer-
antz & J.F. Marshall, Proc. I.R.E., November 1951.


 APPENDIX 1
Listed below are the references from #5 above, reprinted
here to give the reader some idea of the amount of investigative
1 Reliability in Electronic Equipment, F. J. Given, Guided Mis-
sile Committee of AIA, White Sands Proving ground, April
1954.
2 Progress Report on Reliability of Electronic Equipment, Vol. 1
& 2, Research & Development Board, Dept. of Defense,
EL/217 February 1952.
3 Parts Failure Analysis, Report #25, Vitro Corp. of America,
Silver Springs, Maryland, November 1951.
4 Survey of Current Status of Electronic Reliability Program,
Report #RM 1131, The Rand Corp., Santa Monica, California,
August 1954.
5 Reliability Surveys, NEL Reliability Design Handbook, Sec
1.1, U.S. Navy Electronics Laboratory, San Diego, California,
May 1955.
6 Reliability Factors for Ground Electronic Equipment,
McGraw-Hill Book Co. (for Rome Air Development Centre,
Griffiss AFB), New York.
7 Electronic Equipment Failure During Evaluation Test, R.J.
Steelman, NEL Design Handbook, August 1955.
8 Investigation of Electron Tube Reliability in Military Applica-
tions, Aeronautical Radio Inc., General Report #1, Washing-
ton, D.C., January 1954.
9 Study of Electron Tube Life, W.S. Bowie, General Electric Co.,
Owensboro, Kentucky, 1954.
10 Signal Corps Tube Analysis Program, Technical Report #22,
Cornell School of Electrical Engineering, Ithica, N.Y., May
1955.
11 Concerning the Reliability of Vacuum Tubes, M.A. Acheson,
E.M. McElwee, Proc. I.R.E., October 1950.
12 Electron Temperature Problem in Relation to Design for Reli-
ability, Report #51, Vitro Corp. of America, Silver Springs,
Maryland, November 1953.
13 The Background of “Reliability”, T.A. Smith, I.R.E. Transac-
tions on Reliability and Quality Control, September 1956.
14 “Introduction,” NEL Reliability Design Handbook, R. J. Steel-
man, March 4, 1954
15 Heat Dissipation in Electron Tubes, W.D. Campbell, Proceed-
ings of the RETMA Symposium Applied Reliability, Univer-
sity of Southern California, December 1956.
16 The Cooling Effects of a Special Tube Shield, Report #DCA 56-
1161, Aeronautical Radio Inc., Field Engineering Report,
Army Radio Station “WAR” October 1956.
17 Techniques for Reliability Measurements and Predictions,
Based on Field Failure Data, Report #80, Vitro Corp. of Amer-
ica, Silver Springs, Maryland, October 1955.

Page 1.3 11
and developmental work given the problems of tube and conse-
quent, equipment reliability during the 1950s.
18 Vacuum Tube Envelope Temperature Measurements, D.R.
Walin, NEL Report #594, US Navy Electronics Laboratory,
San Diego, California, March 1955.
19 Electron Tube Engineering Data Sheets, Sections 201.1 &
505.1, W. B. Wells, NEL Reliability Design Handbook,
20 Design Manual of Natural Methods of Cooling Electronic
Equipment, J.P. Welsh, Report #HF-845-D-8, Cornell Aero-
nautical Laboratory, Buffalo, N.Y., November 1956.
21 High Temperature and Altitude, Life Evaluation and Pulse
Ratings of Subminiature Vacuum Tubes, Final Engineering
Report, Contract AF33(038)9853, Sylvania Electric Products,
Inc., Emporium, PA.
22 Qualification Testing of Heat Dissipating Electron Tube
Shields, Report #131, United Electrodynamics, div. of United
Geophysical Corp., Pasadena, California, January 1956.
23 Evaluationof Heat Dissipating Electron Tube Shields of Vari-
ous Designs, Report #E-472, United Electrodynamics, div. of
United Geophysical Corp., Pasadena, California, March 1957.
24 Military Specifications, Electron Tubes, MIL-E-1.
25 Evaluation Testing of TR Tube Shield and JAN Bright Tube
Shield, Report #E-154, United Electrodynamics, div of United
Geophysical Corp., Pasadena, California May 1 956.
26 Heat Dissipating Electron Tube Shields, Specification MIL-S-
9372(USAF).
27 Development and Application of a Thermal Diode “Therma-
tron” for Use in Studying Methods of Cooling Electron Tubes,
J.P. Welsh, Report #HF-l 053-D-3, Cornell Aeronautical Labo-
ratory, Buffalo, N.Y., May 1957.
28 Evaluation of Shields for Cooling Miniature Electron Tube
Types, M. Bailer, Report #53-174, Wright Air Development
Centre, Wright-Patterson AFB, Ohio.
29 Interim Test Report on Evaluation of Electron Tube Shields,
United Electrodynamics, div. of United Geophysical Corp.,
Pasadena, California, May 1956.
30 EIA Conference on Electronic Equipment Maintainability,
Paper presented by C. I. Cummings, Jet Propulsion Labora-
tory, California Institute of Technolongy.
31 The Rating of Thermionic Valves for Use Under Abnormal
Ambient Conditions, R.D. Mills & W.W. Wright, British Jour-
nal of Electronics, November 1955.
32 Qualification Test of Heat Dissipating Tube Shields, Report
#E-169, United Electrodynamics, div. of United Geophysical
Corp., Pasadena, California, November 1956.


 L A B R E S U LT S
The outcome of our lab work is shown for vari-
ous tube types under a wide range of operating con-
ditions, both laboratory and real-world.
Please note the truly excellent thermal perfor-
mance of our cooler-chimney-fan arrangement. This
will be incorporated into our upcoming Single
Channel 280W mono-block amplifier which will be
available both in kit form and as a finished product.

BULB TEMPERATURE VS. POSITION FOR THE 9-PIN COOLER

Fig. 9.

With Bare
Cooler Bulb

50 75 100 125 150 175 200 225 250 275° C

122 167 212 257 302 347 392 437 482 527° F

Equipment: Test rig Power – Fil: 2.25W Notes:

Serial number: n/a Power – Plate(s): 1.5W
Tube type: 6DJ8/ECC88 Power – Total: 3.75W
Made by: EI – Yugo Cooler type: 9-pin small signal
Bulb diameter: .80” Cooler length: 1.25”
Bulb area: ≈ 4.25 in.2 Cooler area: ≈ 33.5 in.2

Fig. 10.

With Bare
Cooler Bulb

50 75 100 125 150 175 200 225 250 275° C

122 167 212 257 302 347 392 437 482 527° F

Equipment: Test rig Power – Fil: 4.75W Notes:

Serial number: n/a Power – Plate(s): 14.0W
Tube type: 6BQ5/EL84 Power – Total: 18.75W
Made by: Sylvania Cooler type: 9-pin power
Bulb diameter: .80” Cooler length: 1.75”
Bulb area: ≈ 6.2 in.2 Cooler area: ≈ 47 in.2

Page 1.3 12

 BULB TEMPERATURE VS. POSITION FOR THE STANDARD-FITTING POWER COOLER

Fig.11.

Cooler & Cooler Bare

Forced Air Only Bulb

50 75 100 125 150 175 200 225 250 275° C

122 167 212 257 302 347 392 437 482 527° F

Clear acrylic
plastic chimney

Chassis
Forced air-flow

Equipment: Test rig Power – Fil: 9.5W Notes: PEARL SC 280 cooling-system test rig. Air flow
Serial number: n/a Power – Plate(s): 22.5W ≈ 50ft/ min. through a 2.5” i.d. clear acrylic-plastic
Tube type: EL34 Power – Total: 32.0W chimney
Made by: Tesla Cooler type: 1.25” Std
Bulb diameter: 1.125” Cooler length: 2.5”
Bulb area: ≈ 13.0 in.2 Cooler area: 103 in.2

Fig. 12.

Cooler & Cooler Bare

Forced Air Only Bulb

50 75 100 125 150 175 200 225 250 275° C

122 167 212 257 302 347 392 437 482 527° F

Clear acrylic
plastic chimney

Chassis
Forced air-flow

Equipment: Test rig Power – Fil: 10.0W Notes: PEARL SC 280 cooling-system test rig. Air flow
Serial number: n/a Power – Plate(s): 30.0W ≈ 50ft/ min. through a 3.0” i.d. clear acrylic-plastic
Tube type: KT88 Power – Total: 40.0W chimney
Made by: Chinese: (?) Cooler type: 2.0 ” Std.
Bulb diameter: 2.0 ” max. Cooler length: 2.5”
Bulb area: ≈19 in.2 Cooler area: 150 in.2

Page 1.3 13

Fig. 13.

Bare
Bulb
30W

Cooler Cooler Bare

Only Only Bulb
30W 60W 60W

Cooler &
Forced Air
60W

50 75 100 125 150 175 200 225 250 275° C

122 167 212 257 302 347 392 437 482 527° F

Clear acrylic
plastic chimney

Chassis
Forced air-flow

Equipment: Test rig Power – Fil: 10.0W Notes: The KT90A is shown operating at several dissi-
Serial number: n/a Power – Plate(s): See notes pations. The broken curves show bulb temperatures at
Tube type: KT90A Power – Total: See notes 30W total dissipation while the solid curves show bulb
Made by: EI – Yugo. Cooler type: 1.5” Std. temperatures at 60W total dissipation.
Bulb diameter: 1.5” Cooler length: 2.5”
Bulb area: ≈ 19.5 in.2 Cooler area: 120 in.2

BULB TEMPERATURE VS. POSITION FOR THE CLOSE-FITTING POWER COOLER

Fig. 14.

Cooler Bare
Only Bulb

50 75 100 125 150 175 200 225 250 275° C

122 167 212 257 302 347 392 437 482 527° F

Equipment: Test rig Power – Fil: 9.5W Notes:

Serial number: n/a Power – Plate(s): 22.5W
Tube type: EL34 Power – Total: 32.0W
Made by: Tesla Cooler type: 1.25” CF
Bulb diameter: 1.125” Cooler length: 2.5”
Bulb area: ≈ 13.0 in.2 Cooler area: 95 in.2

Page 1.3 14

Fig. 15.

Cooler Bare
Only Bulb

50 75 100 125 150 175 200 225 250 275° C

122 167 212 257 302 347 392 437 482 527° F

Equipment: Test rig Power – Fil: 10.0W Notes:

Serial number: n/a Power – Plate(s): 30.0W
Tube type: KT88 Power – Total: 40.0W
Made by: Chinese: (?) Cooler type: 2.0” CF
Bulb diameter: 2.0 ” max. Cooler length: 2.5”
Bulb area: ≈19 in.2 Cooler area: 130 in.2

Fig 16.

Cooler Bare
Only Bulb

50 75 100 125 150 175 200 225 250 275° C

122 167 212 257 302 347 392 437 482 527° F

Equipment: Test rig Power – Fil: 10.0W Notes:

Serial number: n/a Power – Plate(s): 50.0W
Tube type: KT90A Power – Total: 60.0W
Made by: Ei – Yugo Cooler type: 1.5” CF
Bulb diameter: 1.5” Cooler length: 2.5”
Bulb area: ≈19 .5 in.2 Cooler area: 115 in.2

Page 1.3 15

APPENDIX 2
ham clinic
CHARLES J. SCHAUERS, F7FE / W6QLV
CQ Magazine, 300 W. 43rd St., N.Y. 36, N.Y.
Some Thoughts On Extending Tube Life
As pointed out in this column sometime ago,
defective tubes cause the majority of failures in elec-
tronic equipment. Contrary to what the advocates of
transistorization may think or say, we are going to
have tubes and tube failures for a good long while yet
The care and feeding of tubes receives a lot of
attention from equipment designers. Whenever pos-
sible, they design circuits so that the tubes are cor-
rectly operated and not allowed to run at high tem-
peratures. But no matter the design approach, there
will always be HEAT to contend with because a tube’s
cathode must get hot in order to operate properly.
Now, when high voltages are applied to plates and
screens, especially those in power tubes, there is yet
more heat to dissipate.
Some years ago, I conducted experiments rela-
tive to the effects of high temperatures on tube oper-
ation and longevity. Although my test setup was
crude and so were my test results, I assured myself
of one thing: tubes do have longer life spans if they
are not subjected to voltage and current overloads
and high temperatures.
Every tube has its own critical operating tem-
perature. This cannot be quickly determined with-
out many hours of environmental testing using spe-
cial equipment. As well, the how and where of a
tube’s mounting can have a lot to do with how long
it will satisfactorily operate.
Heat is a major problem with tubes containing
many elements such as pentode/triodes, double tri-
odes, dual-diode/triodes etc. because the elements
must, of necessity, be so close together. At this point,
tribute must be paid to tube manufacturers for being
able to come up with multi-element tubes that oper-
ate so well over a wide range of temperatures.

Page 1.3 16
Overdriving a tube (with RF , line voltage
surges, etc.) certainly takes its toll, but the one big
“bugaboo” that receives little attention is H-E-A-T,
spelled in large letters!
Sometime ago, I was fortunate to get a copy of
a really fine report titled “Heat Dissipating Electron-
Tube Shields and Their Relation to Tube Life and
Equipment Reliability” prepared by John C.
McAdam of International Electronic Research Corpo-
ration (IERC), 145 West Magnolia Blvd., Burbank,
California. After reading it, I was convinced (as I
know you will be if you read the whole report) that
too little attention has been given to the conserva-
tion of tubes through heat reduction—this being
especially true in ham radio equipment.
Few people realize that the ordinary JAN shield
actually makes a tube run hotter than it would if oper-
ated bare. Take a look at Fig. A to see what I mean!
Taken from Mr. McAdam’s paper, this graph really
shows the difference when an ordinary shield, no
shield and IERC’s special heat dissipating shields are
compared. Note the curves for the TR and B type
shields. Now look at Fig. B and note how much longer
tubes will operate before going sour when properly
designed tube shields are used. Amazing isn’t it?
The findings of various research organizations
indicate that the evolution of gas within a tube due
to elevated temperatures is the principle cause of
tube failure. Other high-temperature-caused fail-
ures are: getter migration, grid emission, glass fail-
ure, inter-electrode leakage, contamination, grid
loading and loss of emission.
Of course, forced air cooling is a solution to the
hot-tube problem and is generally used when possi-
ble to obtain maximum cooling efficiency. But the
mere direction of air over or under a set of tubes is
not always the answer because all tubes do not get
the proper or the same amount of air due in part to
forced mechanical design and circuit layout. Then
too, forced air cooling is not always an expedient
measure in ham equipment nor is it inexpensive!
The shiny surface of the JAN shield reflects heat
back into the tube; nothing better for raising tube
temperature except maybe a nearby, hot transformer.
This is the main reason why all good tube shields are
black inside and out—for heat absorption. Also, the
air space found within the average JAN shield retains
heat, further aggravating the situation.
Referring to Fig. C-1, you will see IERC’s effec-
tive heat-dissipating tube shield. It dissipates the
heat by radiation, conduction and convection. It
grasps the hot tube bulb and distributes the heat
from the hot spot over a large surface area. This way,
it not only reduces the average temperature present
on the tube glass but also greatly reduces the tem-
perature gradient along the surface of the tube.


 420
In Fig. C–2 is shown a retrofit shield developed JAN Shield
by IERC to meet the problem of retaining the old
380
JAN-style base. It merely snaps onto the old type
base. It is capable of reducing the temperature of the

Bulb Temperature in ºC
tube bulb well below the bare-bulb temperature and 340

nearly 100°C below JAN-shield temperatures. This is Bare Bulb

the shield most amateurs can use on the tubes in 300 IERC (TR)
their equipment residing in the old-style JAN bases.
To increase tube life, a tube must be operated 260
IERC (Type B)

properly. This means current and voltages as low as

consistent with proper operation. Neither a tube’s fil-
220
ament nor its cathode appreciate voltage surges so if
it’s possible, use a surge-voltage limiting device: a
variable transformer, Surgistor, etc. Some amateurs 180
20 40 60 80 100 120 140 160 180
(those who can afford it and desire the utmost in sta-
Ambient Temperature in ºC
bility) turn their receivers on and leave them on. In 6005/6AQ5 operating at maximum plate dissipation.
this way, the tubes are not subjected to starting Fig. A. The temperature-effects of running a tube in various ways
surges but they still are affected heat-wise if they are are shown above. Note the undesirable effect of using the brightly
using old-style heat shields. Forced air cooling is fine plated JAN-shield.
if (and this is a big IF) the air can be directed so that
there are no outstanding hot spots. Why cool a tube’s 6005 / 6AQ5W dissipating 16.4W in 100º F ambient
base when its envelope is boiling? If the base is of the 100
heatsink type that’s all well and good but otherwise IERC (Type B)

you’re wasting power. Remember that heat rises so

Survival percentage
80
IERC (TR)
top-mounted ventilation systems are best.
60
Sooner or later, most electronic manufacturers
will get around to giving consideration to the use of
40
the various types of specially constructed heat
shields. I hope that those who make amateur radio 20
Bare Bulb

equipment will give these items special attention. JAN Shield

Even with properly operated and cooled tubes we 0
0 500 1000 1500 2000 2500
will still have replacement requirements, so tube Operating time in hours
manufacturers can take heart. They may have to
Fig. B. Tube life as a function of the sort of heat-dissipation mech-
wait just a little longer for replacement orders and anism employed. Note, once again, the dismal performance of the
tube prices might rise slightly, but even the best brightly plated JAN-shield.
cared for tubes—like humans—do not last forever.

No one can ever say unchallenged—in my pres-

ence anyway—that American hams are not person-
ally generous or are disinterested in international 1 2
goodwill! The response to my appeal to send foreign
hams your old issues of CQ, QST, WRA, CRA, QSO
etc. is heart warming! Bravo!
I am sure that the recipients of your magazines
and call books will not look on your gifts as charity
but rather as your contribution to solidifying inter-
national ham friendship. There will surely be some
reciprocation and I sincerely hope that you will take
the time out to thank the donor, whoever he may be.
Some foreign hams will be surprised and won-
der what brought on this sudden generosity. All I can
say in reply, is “72” to you: “Peace and Friendship in
Freedom”.
Note: I have taken substantial editorial liberties Fig. C. – 1 IERC’s Type A heat dissipating tube shield. – 2 The
with this article. While the tone and content remain, it IERC retrofit type shield. While not as effective as the Type A, it will
now reads rather more clearly. bp. fit on the JAN-type base.

Page 1.3 17

APPENDIX 3
Fig. 8. These data are a re-drawing of an original plate from a Tung-Sol Tube-Data Manual.
† Emphasis added.

TUNG-SOL
BULB TEMPERATURE CURVES
One of the most important factors affecting the useful life of electron tubes is the temperature at which cer-
tain parts are required to operate.† In the past this has been controlled largely by electrode dissipation rat-
ings. Recently a few, but now an increasing number of tube types have been rated for maximum allowable
bulb temperature in addition to these dissipation ratings. The following curves relate the approximate “hot-
spot” bulb temperature to the total dissipation (including heater power) for various sizes of bulbs under
arbitrary reference conditions. Therefore, if the dissipation is known, these curves may be used to estimate
whether or not the bulb temperature rating would be exceeded under such conditions. However, sufficient
departure from these conditions would require actual temperature measurement.
The curves may also be used to find the approximate dissipation indirectly from bulb temperature when
complex non-linear voltages and currents, such as are frequently encountered in radar, pulse and televi-
sion service, make measurement by conventional direct methods very difficult if not impossible. Data for
the curves were taken by applying DC voltages to the indicated representative types in a 5″ x 5″ x 7″ enclo-
sure, and measuring the hottest bulb temperature with an iron-constantan thermocouple made of .003″
wire. This “hot-spot” is usually found two-thirds to three-quarters of the way up the plate structure near the
place where the plate is closest to the glass. Any bulb temperature measurement should be made with a
thermocouple that is made of very fine wire. In addition, great care must also be taken to minimize con-
vection cooling, and to allow sufficient time to obtain a stable reading.

350

BULB TEMPERATURE
VS
DISSIPATION
BULB TEMPERATURE AT HOTTEST POINT - °C.

300 LL
OATED

TA
1/2 M D
T-
5 IU ZE
ED NI T-14
5) 6) S
VER C

M O
Q /2 RB R 1 (6 G Y T-16
6A CA EA T-1 S
CL
61

6) )
G6G
(

-9 T-9 AR
T-

(6B
T-3 SIL

T ) (6
7)

T
T) 6G
68

200 6G (6V
T

(5
OR

V
(6
3

SH
(5636)
6) T-

/2
- 51
(563

)T
K5
(6A

150

100

NOTE : ALL DATA TAKEN AT THE HOTTEST SPOT ON

THE BULB AS FOUND AT RATED DISSIPATION WITH
THE TUBE IN A VERTICAL POSITION.

A BOX 5" SQUARE BY 7" HIGH ENCLOSED THE TUBE

50 TO MINIMIZE TEMPERATURE VARIATIONS DUE TO
ROOM AIR-CURRENTS

0
0 5 10 15 20 25 30 35
TOTAL INTERNAL DISSIPATION INCLUDING HEATER POWER - WATTS

TUNG-SOL ELECTRIC, Inc. ELECTRON TUBE DIVISION NEWARK, NEW JERSEY, USA AUGUST 1, 1953 PLATE 3296

Page 1.3 18

 APPENDIX 4.
Product reviews and comments.
The following is a review commissioned by a
British magazine, the rights to it were acquired by
our UK distributor:
Russ Andrews Turntable Accessories
Edgebank House
Skelsmergh, Kendal,
Cumbria, England
LA8 9AS
Reprinted with permission.
PEARL Valve Coolers
Review by Roy Gregory
PEARL valve coolers are concertina, blackened cop-
per sleeves that slip over a valve’s glass bottle, and are
clamped against it with thick rubber bands. They come
in a whole range of sizes to fit virtually any available
valve and fitting is fairly easy depending on the shape
of the valve in question, 2A3s and old-style bulged and
tapered 6550s are the worst.
The theory is simple. The black copper sleeves act as
chimneys, conducting heat away from the tubes and
thence to the air. This reduces operating temperatures
and increases valve life. The clamped sleeve also damps
the bottle and reduces microphony.
So much for theory. In practice your main problem
is going to be space. The ‘fins’ that the coolers add to a
valve increase its circumference substantially. Where
valves are packed together PEARL suggest you don’t use
the rubber bands, and interlock the coolers, like cogs. I
disagree. No, let me put that stronger. Don’t do so
under any circumstances. Now let me explain.
The PEARL coolers are certainly efficient chimneys.
If you don’t believe me just touch one. Remove the
band(s) and you reduce that efficiency, but that’s not the
real problem. Long term testing will show how effective
they are in prolonging valve life (and as I write a P35
MKII kindly donated by Beard is driving a dummy load,
its valves variously cooled and uncooled) but their real
value lies in the sonic difference they make as dampers.
Add the PEARL coolers to an amp and you’ll notice an
immediate improvement in midrange clarity and focus.
The top end gains sweetness and purity, losing grain
and fog, and individual instruments are easy to pick
out. Listen to a cymbal or triangle decay. More detail,
clarity and a measured fading of the sound. Listening
PEARL-less the same strike will be lost in glare and fuzz.
A well worthwhile upgrade then.
The problem is that the coolers sans elastic bands

Page 1.3 19
can sound worse than no coolers at all, creating fuzz
and grain right across the upper mid. I can only assume
that they’re picking up airborne vibration and rattling
against the valves. Whatever the explanation is, I didn’t
like the result, so check your spacings with RATA before
ordering (PEARL supply excellent data sheets with all
their products, so confusion should be kept to a mini-
mum). Oh, they look pretty post industrial too. I kind of
like it now I’ve got used to it, but if your sole reason for
owning valve equipment is to bask in its seductive glow,
then these are not for you.
My success with the coolers encouraged me to try
something a little more demanding, PEARL’S replace-
ment valve bases. These come in a variety of forms with
ultra high quality gold plated contacts and chassis or
PCB mounts. I went for a quartet of the most elaborate
version in which each socket floats on a pad of Sor-
bothane, connected to the circuit by fine, Litz wires.
The long suffering P35 was pressed into service and
after an hour or two I had a single pair of suspended
output tubes per side, allowing direct comparison with
the stock ceramic bases normally fitted.
The P35 is getting a bit long in the tooth but it
always sounded better on single pairs of output tubes.
Comparing the suspended valve bases to the normal
ones revealed a huge, almost shocking difference. The
coolers had improved things, but in largely cosmetic
ways. The valve bases wrought a fundamental change.
Prior to fitting them the P35 had sounded grainy and
disjointed; nice enough but not very interesting. Using
suspended output valves gave the music an immediate
sense of cohesiveness and authority. It was almost as if
all the information the amp had been passing was sud-
denly rearranged in the proper order. The stereo per-
formance opened out; the bass went deeper and
bounced along in time with the rest of the music; feet
began to tap and the performance made sense.
Somewhat bemused I ran the comparison past
friends and visitors, who were all similarly gobs-
macked †, so I guess I’m not imagining it. Either way,
the valve bases turned an also ran into a very real con-
tender. They aren’t cheap at around £25 each (!) but
they made way more than £100 of difference to the P35.
They also narrowed the gap between Gold Dragon
EL84s and Sovteks, compensating in part for the valve
quality. So far I’ve only tried them on output tubes, but
low level valves (the phono stages in valve pre-amps!)
are crying out for them. Beware of horizontal mount-
ing, though, as the Sorbothane may well “flop”, it’s so
spongy.††
The P35 in its final form, PEARL cooled-and-sus-
†A colloquialism that roughly translates to, “Slapped up the side of the head”.
†† The horizontal mounting Iso-sockets don’t use Sorbothane for exactly this
reason. “Non-flopping” silicone foam is used instead, see Audio Note 6.1, Figs.
4, 5 & 7. bp.


pended with Sovteks in place, is an entirely different
kettle of fish from the amp that arrived on my doorstep.
Fitting the valve bases requires patience but unless you
need bigger holes in your top plate, I can’t see it giving
any problems. The results are more than worth the cost
and bother (PEARL make cheaper and less complex ver-
sions as well). Long term results on the coolers will have
to wait a while, but we should get empirical evidence of
the effectiveness (or lack of it). Meanwhile, they are the
most effective tube dampers I’ve found, outperforming
Sorbothane or Sicomin varieties, costing less and work-
ing with almost any valve! Those worried by their metal
construction should remember that copper is non-mag-
netic and the preferred chassis material of the fine
Japanese Audio Crazies. The PEARL products are per-
fect for those who want to improve their tube equip-
ment, vintage or current, offering better sound from a
straight substitution, and the pride of doing it yourself.
A totally rebuilt Stereo 20 is definitely in the cards, Chez
Gregory.
The following appeared in the August 1994 issue of:
Hi-Fi News & Record Review
published by:
Link House Magazines Ltd.,
Dingwall Avenue, Croydon,
CR9 2TA UK
For subscription information, US & Canadian readers
may call 1 800 688 6247 toll free.
Reprinted with permission
BIRTH
OF THE
COOL
PEARL TUBE
COOLERS are
a worthy
alternative to
Audioquest’s
tube dampers !

Page 1.3 20
or whatever reason, the tube damper from Audio-
F quest is out of production, so it’s time to find a
replacement. And, after ruling out a few due to crimi-
nal pricing, we’ve settled on one which is more than
just a bargain. It’s multi-purpose.
Admittedly, the Pearl Tube Cooler is a cooler first
and a damper second. The damping is almost inciden-
tal, a by-product of any device which clamps to a valve.
And even though damping microphony is not is pri-
mary function, the Pearl Tube Cooler works almost as
well as any in my arsenal. Boring and time-consuming
though it was, I A B/C’d (in a Croft pre-amp) the Pearls
against sorbothane, Kevlar and other types and found
it ‘almost, but not quite’ provided the same kind of
immunity from treble smearing and bass sogginess
available from the Audioquest. But since the Pearl has
another function (and costs a pittance), I’m not about
to sit shiva for the Audioquest device.
Tube cooling is desirable for one main reason
which should prove appealing to the more venal among
you: cool-running tubes last longer. According to its
maker, the use of a Pearl Tube Cooler should double a
tube’s life, in some cases extending it by a factor of five.
The Cooler itself is a clever design made from
blackened sheet copper fashioned into a pleated sleeve.
Because of its springiness, it’s impossible for the Cooler
to be ‘too tight’ should you order the wrong size. If a
Cooler is fitted too large a valve, the pleats simply
spread out a bit more than normal. But you shouldn’t
have a problem because six types are available, for
everything from miniature valves like EL84s up to
power tubes like EL34s and KT88s, and for tubes with
lots of space around them or those nestling close
together. With the exception of amplifiers like the Rad-
ford STA25, where the tubes are as close as this, most
amps can be accommodated; the minimum space
allowed between valves is just over one-half inch. The
sleeves slip over the valves; those for more widely-
spaced tubes come supplied with heat resistant rubber
bands to keep them in place, while close-fitting types
can be ‘interwoven’.
I used the small, standard-fit Pearls with the
GRAAF GM200, the Unison Research Simply Two and
the latest Croft Micro. I found that the Pearls quickly
grew scalding hot, conducting heat away from the tube
by adding a large amount of heat sink surface area.
Additionally, the Pearl Cooler’s shape creates a chim-
ney effect, so convection takes the heat away as well.
There’s no way I can tell you about the life-extend-
ing properties of these devices after only a fortnight’s
usage, but I can attest to their worth as replacements
for the Audioquest dampers.
Bottom line? I love ‘em because they look so
obscenely macho.
Note: Upon applying for Pearl Tube Coolers from the
Accessory Club, you will receive a special order form
with a measuring template and full instructions to help
you select the correct size and type. Ken Kessler


 The following appeared in the May 1993 issue of:
Stereophile
published by:
Stereophile
208 Delgado
Santa Fe, NM
87501
For subscription information, US & Canadian readers
may call 1 800 238 2626 toll free.
Reprinted with permission
TUBE AID 2
Jack English looks at ways of improving tube
performance & prolonging tube life
AudioQuest Sorbothane Vacuum Tube Dampers. Price:
$15/set of four. Manufacturer AudioQuest, P.O. Box 3060,
San Clemente, CA 92672. Tel: (714) 498-2770. Fax: (714)
498-5112.
PEARL Tube Coolers. Price: $11.25 each for Power-Tube
Coolers, $5.50 each for Small-Signal Tube Coolers. Approx-
imate number of dealers: 15. Manufacturer: Perkins Electro-
Acoustic Research Laboratory, 2510 19th St. S.W., Calgary,
Alberta T2T 4X3, Canada. Tel: (403) 244-4434. Fax: (403)
244-9026.
Ensemble Tubesox. Price: $48/pair. US Distributor Ensemble
USA, 111 Oak Place, Hendersonville, TN 37075. Tel: (615)
8240047. Fax: (615) 824-3470.
Many of us love the sound of tubes. Unfortu-
nately, living with tubes often means living in fear—
fear of “tube mortality”; fear that they will fail when
we turn something on; fear they will begin to deteri-
orate just before all our friends come over to listen;
fear of the ubiquitous microphony appearing at the
worst possible moment.
I keep an ample supply of replacement tubes on
hand for all these inevitable occurrences, but while
this is an effective strategy, it is an incomplete—and
expensive—one. Instead of simply accepting frequent
tube failure as part of the price of living with the cher-
ished fire bottles, we can do something about it.
The internal workings of tubes are mechanical;
in use, they vibrate and resonate. When the reso-
nance can be heard, the phenomenon is called
microphony, one of the most common and frustrat-
ing aspects of tube behavior.

Page 1.3 21
One solution readily available for this particular
problem is the Sorbothane Tube Damper from
AudioQuest, reviewed by Dick Olsher in February,
Vol. 16 No. 2, p.176. Each Damper looks something
like a very wide, thick black rubber washer. The
Damper can easily be installed even if the tube is
already in place: simply ease it over the glass casing
of the tube, much like a fat rubber band. AQ sug-
gests that two Dampers be located near the top and
bottom of the tube to cause the least interference
with radiant cooling. However, I have found the best
damping to occur with the Dampers placed near the
center of the tube.
One tube with a reputation for microphonics is
the 6DJ8; for years I have used Tube Dampers on
every 6DJ8 that has found its way into my system. I
also use Dampers on all phono-stage tubes. While I
may have been unusually lucky, I have not experi-
enced the host of microphony problems that many
audiophiles complain of. I attribute this to the effec-
tiveness of the Dampers. Where I have had prob-
lems, the Tube Dampers have either helped or virtu-
ally eliminated the problem. While I can’t absolutely
say that the devices prevent microphony, my own
experience has convinced me to continue using
them religiously on all new tubes. Their only limita-
tion is that of size: Tube Dampers are available only
for small tubes (eg, 6DJ8s, 12AX7s). The second
problem with Tube Dampers is that they are gener-
ally not reusable. In some cases they become
bonded to the glass casing of the tube. In other
cases, the heat of the tube actually melts the Sor-
bothane. Given Sorbothane’s low melting point,
Dampers cannot be used on power-supply or output
tubes (eg, 6550s, KT88s), which generate consider-
ably more heat.
Heat is the most common tube bugaboo. Here
again, a number of useful solutions are available.
One is the Tube Cooler from the Perkins Electro-
Acoustic Research Laboratory (P EARL ). These
devices are claimed to significantly reduce the oper-
ating temperature of the tube’s glass envelope, thus
dramatically extending tube life. Extensive docu-
mentation to support these claims is available
directly from Bill Perkins at PEARL in the form of a
series of papers called “Audio Notes.” Unlike Tube
Dampers, Tube Coolers are available in multiple
sizes to fit most tubes. The most popular sizes are
the Small-Signal Coolers (for 6DJ8s, 12AX7s, etc.)
and Power-Tube Coolers (for 6550s, KT88s, etc.).
Since tubes from different manufacturers are often
of different sizes, specifically sized Coolers are avail-
able for precise fits. Each Cooler is formed from
blackened, foil into a radial fin arrangement looking
much like the heatsinks found on the exteriors of
amplifiers. The tube is inserted into the Cooler and


held firmly in place by high-temperature-resistant
elastomeric bands (two silicone O-rings per Cooler).
End-on, the Cooler looks like a gear with the tube in
the center and fins radiating outward, perpendicular
to the tube’s circumference. Being made of metal,
the Coolers do introduce unpredictable changes to
internal electrostatic fields. In some applications,
users have reported negative sonic effects. In part to
remedy these concerns, P EARL offers extensive
installation advice. Each Cooler can be grounded
three ways: with a simple piece of wire, using resis-
tors at both ends of the ground wire, or using resis-
tors and a small capacitor. Once grounded, the Cool-
ers should continue to achieve their intended pur-
pose of extending tube life with no audible draw-
backs. As they are made of metal, it’s possible for the
Coolers themselves to vibrate. The dual O-rings, in
addition to holding the Coolers firmly in place, act
as dampers. I tried Coolers in both the CAT SL-1 Sig-
nature and the Melos 333 line-level preamp/phono
stage combination. I heard no obvious mechanical
problems, nor was I aware of any odd effects which
PEARL TUBE COOLERS
could have been attributed to altered internal elec-
trostatic fields. There were no sonic differences with
the Coolers in place. Had there been, the Coolers
could have been made inaudible via one of the sug-
gested grounding approaches described above.
The newest entrant in the improved perfor-
mance/tube-life extension category is Tubesox from
Ensemble (also reviewed by Dick Olsher in Febru-
ary, Vol. 16 No. 2 , p. 176 ). What makes Tubesox
unique is that they deal with microphony and heat.
Each Tubesox is approximately 1 1⁄4 ” long and
appears to be made of straw tightly interwoven with
copper strands. The devices look much like “Chinese

Page 1.3 22
handcuffs”: stick a finger in each end, try to pull
them out, and the weave tightens, gripping your fin-
gers. The woven construction damps the tube while
the copper wire is intended to dissipate heat from
the glass casing. Tubesox look enough like straw
that I was concerned that they might be a fire haz-
ard. I mentioned this to George Bischoff of Melos,
who suggested trying to ignite one. With ashtray in
place and matches in hand, I lit one of the Sox. It
scorched with extensive exposure to the flame, but
would not catch fire. As it turns out, the Sox are
made of Kevlar and are definitely not flammable.
Whew! Like the Dampers, Tubesox are available
only for small tubes like 12AX7s and 6DJ8s; unlike
Dampers, they are small enough to fit anywhere the
tubes themselves will fit—each thin, glove-like Sox
takes up far less room than a Damper or Cooler. In
really tight spaces, the Tubesox might be the only
possible choice among the products in this article.
Like Dampers, Tubesox fit tightly around the glass
casing of each tube, acting to restrict mechanical
vibration and thus reduce unwanted microphony.
ENSEMBLE TUBESOX
Though both Coolers and Tubesox use copper as
a means of dissipating heat, it is unlikely that the
Tubesox would be nearly as effective at this, as the
Coolers are much larger and made entirely of copper.
I did not conduct experiments to verify this assump-
tion, however. On the other hand, as the amount of
metal in the Tubesox is minimal compared to that in
the Coolers, the former are not as likely to create
unpredictable effects in the tubes’ internal electrosta-
tic fields, nor are they likely to resonate. While all
three of these products can help improve tube per-
formance and/or prolong tube life, there are still
some very basic things that should be done with all


 MORE ON TUBESOX
After digesting my February review of the
Tubesox (Vol. 16, No. 2. pg. 176), Ensemble’s Urs
Wagner called me to discuss a couple of specific
points. First, concerning my finding that one
size does not fit all—I was wrong. He pointed out
that the sock diameter may be altered a few mil-
limeters by stretching or squeezing it. Ah-ha! I
tried it. Indeed, via this simple expedient I can
securely fit Tubesox to any preamp tube I’ve got
on hand.
Second, concerning the temperature mea-
surements: while I found that the Tubesox
caused a slow temperature-rise of several
degrees at the top of the glass envelope, Wagner’s
data at a point under the Tubesox (closer to the
tube equipment. First, proper ventilation is
essential—higher temperatures definitely shorten
tube life. Make sure there is adequate space around,
and especially above, your equipment for proper ven-
tilation; if possible, use low-noise fans. If the equip-
ment has special turn-on circuitry, use it. If not, do
not immediately put a signal through the gear. Turn
it on and let it warm up for 20 or 30 minutes before
using it. If the equipment has turn-off features (eg, a
higher, noisier fan speed), use them when turning the
equipment off. Another alternative comes in the
form of specialized pieces of equipment—like the
SimplyPhysics Variac—that allow current to be
turned up gradually. We don’t have to passively
accept microphony and premature tube failure. Tube
Dampers, Tube Coolers, and Tubesox offer ways to
improve tube performance and longevity. Dampers
help with microphony; Coolers dramatically extend
tube life; and Tubesox help with microphony and
offer some benefit for extending tube life.
MANUFACTURER'S COMMENT
I appreciate Jack English’s efforts in the evalua-
tion of the PEARL, INC. line of tube coolers, his com-
ments about living “In Fear of Tube Mortality” really
hit home around here. I have done a lot of investiga-
tive work to discover—and develop products that
eliminate—several causes of the unreliability often
experienced with late 20th-century, high-dollar, tube-
type audio equipment. It turns out that the very relia-
bility issues facing today’s tube lovers were thor-
oughly researched and resolved 30 years ago. The old
research work is not lost and can be brought to light
once again for the price of a few Sunday afternoons’

Page 1.3 23
plate) yielded a small temperature drop. What
this appears to mean is that, although the aver-
age temperature of the tube rises with the Tube-
sox in place, the sock is effective in smoothing
out hot spots on the surface of the tube. If this is
the case (I haven’t attempted as yet to duplicate
Ensemble’s measurements), the Tubesox could
arguably be said to enhance tube life—but only
if a small reduction in hot-spot intensity is more
important than average tube temperature. How-
ever, let me emphasize again that the essential
reason for using Tubesox is sonic, not thermal.
Tubesox can elevate an ordinary-sounding pre-
amp tube to the exalted status of a premium
tube. —Dick Olsher.
digging (alright… more than a few) through any of
the many good technical libraries around the country.
The Audio Note Series is a distillation of years of
library archaeology and contains a lot of information
that is useful to anyone who owns or uses tube equip-
ment, the cost is $12.00US.
The reliability problems seen in present day
tube gear are, with most American and European
tubes, not the fault of tubes themselves but, those of
the circuitry in which they are placed.
This is a time when many of the broadly experi-
enced and deeply knowledgeable old-timers are
either retired or no longer among us. There seems to
be an attitude among present-day designers that on
account of the loss of the “good old guys” from
active audio work, there is unbridgable gap in the
knowledge that can brought to bear on the design of
tube circuitry. It is consequently far from uncom-
mon to see glaring errors within tube circuitry;
errors that cause tubes to run far outside their spec-
ified range of operation. Three of the most com-
monly seen mistakes are:
• the violation of the heater-to-cathode voltage
rating
• the use of power supplies that apply full B+ to
tubes whose cathodes have not been given
time to get to operating temperature
• the packing together of power tubes like sar-
dines in a tin.
The heater/cathode combination is the part seen
glowing red-orange inside a correctly functioning
tube. The heater is fitted inside a coated, nickel tube
that it is the cathode. It is electrically insulated from
the cathode by a thin and somewhat physically frag-


ile ceramic coating that has very definitely defined
voltage hold-off capabilities, typically on the order of
±100VDC. In typical tube circuitry, the heater supply
is held at signal-ground potential ie. 0V. If the cath-
ode is tied to ground by a resistor across which there
is a voltage drop of 100VDC or more then the fila-
ment supply must be biased to some value that is
within 100VDC of the cathode potential. The heater-
to-cathode insulation can leak or, at worst, develop
an outright short thereby connecting the cathode to
the grounded heater. In the former situation the out-
come can be some very peculiar intermittent noises
while the latter case can result in the circuit going
south with an unbridled vengeance.
Cathodes are damaged by the application of full
B+ voltage to tubes that have not been given time
come up to full operating temperature. It is neither
complex nor expensive to build a slow-starting B+
supply, yet such circuitry is practically never seen
today whereas 30 years ago tube rectification pro-
vided a simple, effective solution. [Guitar amplifiers
from the ‘60s with solid-state rectification invariably
switched the filament and B+ supplies separately and in
that order for the same reason.—Ed.]
From about 1940 to 1965 the British and US
Armed Forces and the airline/aerospace industry
spent tens of millions researching the causes of and
remedies for equipment failure. Fundamentally, the

Page 1.3 24
results were that tubes must be kept as cool as possi-
ble and either specially built to withstand shock and
vibration or mounted in ways that provide mechani-
cal isolation from those forces. In study after study it
was found that tube-life could be at least doubled and
often extended many times by the application of rela-
tively simple measures to remedy these problems.
Given good quality tubes, correct circuitry and
careful thermal and vibration management, there is
no reason why tube-type audio gear cannot give thou-
sands of hours of failure-free service.
One of the best examples of tube reliability is
their use in the original trans-oceanic telephone
lines. Scores of tube-type repeater amplifiers were
permanently installed within the several-thousand-
mile length of lines that were laid in the deep water
between continents. Admittedly, the tubes were the
best that Bell Labs and Western Electric could build
but the fact remains that they ran and ran reliably
for years on end with no service whatever.
Ma Bell didn’t hire Captain Nemo to swap out a
bunch of tubes every time Roosevelt wanted to talk
to Churchill.
BILL PERKINS
PEARL, INC.
CALGARY, CANADA


 P EARL Valve Coolers
I’ve been using PEARL Valve
Coolers on my Jadis JA30s since I
first came across them, getting on
for ten years ago. For the uniniti-
ated, a P EARL cooler is a con-
certina sleeve constructed from
sheet-copper which slips over an
audio valve, acting as a heat sink.
The copper fins are perforated
with horizontal slots to improve
airflow, and the whole thing acts
as an extremely efficient chimney
(If you want to know how efficient
then PEARL can supply you with
reams of data on the subject. The
theory of course, is that the sleeve
allows the valve to run at a lower
temperature, leading to longer
tube life and greater reliability But
there’s more to it than just that.
Along with the cooler you get
a pair of stout rubber O-rings
which are used to clamp it in
place (Useful Tip: put the rings
over the cooler and then slip it
over the valve—it’s a lot easier!).
The sleeve now also acts as a
damper, reducing microphony
and improving sound quality And
they work. You should notice a
very real improvement in trans-
parency focus, low level informa-
tion and dynamic range once
you’ve installed the coolers. Do
they improve the reliability and
longevity of your valves? Well,
nothing has blown up and I’m still
running the same Platinum Grade
Gold Aero ECC82 s and 83 s as I
was when the coolers were first
installed. And they still sound
much better than the stock items
currently supplied with the
JA100s. Given the price, that’s a
real relief. As regards the output
tubes that’s harder to gauge, but
the EL34s seem to be lasting a lot
longer than I’d expect as well.
Downsides? Apart from the
by Roy Gregory
fact that they hide
the pretty glowing
bottles, not much
that I can mention.
Me, I kind of like the
Fritz Lang look. You
need to check the
space around your
valves to make sure
that there’s enough
room for the cool-
ers. PEARL suggest
that you can inter-
lock the fins, rather
like cogs. Don’t. To
do so you need to
dispense with the O-
rings, and the cool-
ers end up rattling
against the tube,
adding to the
microphony rather
than damping it. In
short, it’ll make your
amp sound worse. [See our com- tion suggests that they are every
ments below. PEARL] Also, if you’ve bit as effective on pre-amps. The
non-parallel sided tubes then the O- “improvement” comes in the
rings have a tendency to roll down shape of a woven carbon fibre
to the narrow end under their own sleeve that fits between the cooler
tension. The new Sovtek KT88 and the valve, improving the
springs to mind. Finally, you need mechanical coupling. [Aug, 2010:
to order the right sizes. PEARL pro- The carbon fiber sleeves were
duce coolers in a whole host of dropped a number of years ago to
dimensions, so it’s worth taking a be replaced with a high metallic
little trouble over getting things content powder coat finish that
right. That aside it’s a case of fit and provides considerable damping of
forget. the cooler and useful damping of
In fact, it’s so easy to forget the tube’s resonant internal struc-
about the P EARL s that the only tures.. With time, pressure and tem-
thing that prompted me to write perature this coating slightly flows
this piece, was the arrival through to conform to the inevitable irregu-
the post of a quartet of latities of the glass envelope,
“improved” models for ECC83 thereby increasing both thermal
sized valves. As this includes just and mechanical coupling. PEARL]
about every low-level signal or dri- Replacing the old-style coolers
ver valve used in audio electron- on the 30s with the be-stockinged
ics, that makes it very interesting variety produced a further subtle,
indeed. I only use the coolers on but worthwhile benefit. Subtle in
my power amps, but experimenta- terms of putting your finger on it,

Page 1.3 25

far from subtle in terms of simply
identifying it. Greater transparency
and an increase in the air and
space around and between per-
formers, produced a noticeable
increase in presence. At the same
time, a further improvement in low
level dynamics made instruments
more vibrant and real. How big
was the margin of difference? You
didn’t need to resort to ABA com-
parisons. The benefits were
clearly (!) apparent, to the extent
that Victoria commented on the
improvement when she arrived
home that evening. They’re also the
kind of changes that can be readily
masked in a system wanting in the
areas of resolution and low level
dynamics, so if you fit PEARL cool-
ers and don’t hear a difference then
either your valves are immune to
vibration, or it’s time to take a close
look at your system and its set-up.
And remember, I got these results
simply upgrading half of the cool-
ers on otherwise fully “PEARL-d”
amps. The comparison between
Issue 5 - Feb/March 2000 of HiFi +
un-cooled tubes and the same
valves with their modesty intact,
running in the Graaf 13.5B, was
absolutely huge, and out of all pro-
portion with the costs involved.
The carbon socks are available
separately but only come in ECC83
or EL84 sizes. Presumably although
carbon fibre is a good conductor of
heat (remember all those bright red
brake shoes on Formula I cars),
using it with the larger power tubes
compromises the cooling effect.
[These are no longer supplied, May
2008. PEARL] Personally I’m much
more interested in the immediate
sonic benefits than the possible
increase in valve life, so I’m keen to
try socking my EL34s too. Perhaps
PEARL will give us the choice? [See
our previous comments. PEARL]
PEARL Coolers are one of the
most cost effective and sensible
upgrades on offer to valve users.
After all, if you don’t like the
results you simply take them off.
At a cost of £7.00 each for the
small size, including sock, and
The foregoing review appeared in
Contact Info:
Hi-Fi + Editorial
PO Box 1716,
Fordingbridge, Hants.,
SP6 1SL, United Kingdom.
Ph. 44 0 1425 656004
Fx. 44 0 1425 656046
http://www.hifiplus.com

Page 1.3 26

£9.50 for the EL84 size or larger
(“Gee! Do you s’pose they do them
for 845s?”), [Yes they do, see our site
here. PEARL] you need to be run-
ning a power station before the
cost becomes prohibitive. If you
already use the standard coolers,
you can get socks for around £2.50
each. And yes I did try the socks
on their own, and no they aren’t
nearly as good as the sock/cooler
combination. I love these things.
Cost effective and no-nonsense,
they even have empirical data to
back them up. Unlike too many
things in hi-fi, buying these is an
absolute no-brainer.
Distributed by:
MOTH MARKETING
Tel. (44)(0)1234-741152
Fax. (44)(0)1234-742028
e-mail.
moth@britishaudio.co.uk
Net. www.britishaudio.co.uk
APPENDIX 5.
An ad from the mid-1950s . . . and no, we didn’t copy this.

Page 1.3 27

♦ Verso Filler Page ♦
 6AK5
6J6

12AU7
6AQ5
6AR6
o
... 6SN7

�.. 12AT7

... 2C51
a:
6AG7
t3 w
Q.
CD >- 6C4
:> ....
t- 6J4
w
ID
:> 6V6
.. .... '
z 5R4GY

6AC7
� 6AL5

CABANISS FORT WAR NORFOLK CARS- EUCOM MACDILL ALL D 200 400 600 800 1,000
BLISS waL BASES NUMBER � ;UBES RETURNED

FIG. I-Tubea retumed fall Into four defect clasaea: mechanical. FIG. 2-Tube types 6AK5 and 6J6 lead the twenty tube type.
electrical. miscellaneous Includinq breakaqe and no defect rankinq hiqhest in l o tal number of relums for first quarter.
1953

Military Reliability of

Study of over 150,000 electron tubes used in military equip ment reveals typ es most
likely to fail and discusses common kinds of failure. Tells what equipment designers
and tube manufacturers can do to increase electron-tube reliability

By E. R. JERVIS and J. SWAUGER

Aeronautical Radio, rllc.
Washington. D. C,

EVERAL
S
long-term programs jected tubes from military equip­ includes data from over 44,000
have been launched by the ment. The tubes, with available tubes .:ollected from 44 general
armed services with a view to im­ failure data, are forwarlled to e qu i p ment types.
proving the rel iabil i ty of more than Washington and subjected to both
Types of Defects
30 tube types. These types will be e ngineer i ng and statistic al analyses.
usable in more than i of all sockets In two years, 88,500 tubes re­ Tube returns have been classified
in present-day military equipment. moved as defective by military in fl)ur general defect categories,
The improved types are being de ­ technicians have been collected, Electrical defects include tuhes re­
veloped to reach a goal of 95-percent Over 62,000 tubes have bep.n in­ jected because of deterio ration of
survival in 1,000 hours operation in stalled in controlled tests. Tubes performance caused by faulty pro­
contrast with SO-percent in 500 under test include over 14,000 of cessi ng during tube manufacture,
hours for presently available tubes, the 22 improved tube types as well end of normal tube l ife and environ ­
A large-scale surveillance project as many thousan ds of their JAN mental 01' ope r atin g conditions.
is under way to determine the prototypes. Mechanical defects are physi c al de­
status of military electron tube Most of the information to be fects in the structure caused by
reliability before and after use of disc ussed concerns performance of enviro nmen tal conditions or weak­
improved tube types, to discover JAN tube types since the improved nesses of tube c o nst ruct i on. l'Iiscel­
causes uf unsatisfactory p..;rform­ versions have been used only in ianeolls defects include broken glass,
ance and to suggest measures for limited numbers during the last noise, microphonism and defects
improving reliability . Field stations two ye a rs . This article covers from caused by unusual environmental
at eigh t military bases collect re- Sept. 1951 through March 1953 and co n ditions or mishandling by main-

130 April, 1954 - ELECTRONICS

 12SK7, 6AQ5, 6AU6 and 2C51.
�IOO r-----'
With the E'xception of the 12SK7
�
t; these types are of relatively recent
a:
III 57.1 design and critical a ppl icati o n .
'" t----tl._--t. IHI+------- ---------! Tubes in the third category­
�:>O
types 6AK5, 6J6, 12AT7, 6AS7 and
�
12AU7-show large returns due
�:I both to large socket populalions and
� 0�IL95��
1 .w�����W.LI9
- 5L2��--�
� -L-��--�
��� � 5 3
to abnormal removal from l'ome of
TIME IN 4-:WEEK INTERV.:ILS
(A) the sockets in use. The military
services ha\"e already begun i 111-
provement p roj e cts on 17 of the

J09B
20 types listed in Fig. 2.
105.0 _�I ..
-==..::-t
-- ' .. _-

1
__ - -

�.:..o_
------- _.

�--
__
-- . -
Type 6AK5

I
The only outstanding weakness

1951
I I1952 1953
revealed in the 6AK5 is th(> heater.
Table I gives a q ua n� itat h e com­ '

TIMF IN 4·WEEK INTERVALS parison of the reliabi lity of the

CB)
6AK5 and its improved versions,
FIG. 3-Ret um a of SARS's from power supply of radar bomblnq system (A) and types 5654 and 6096. Comparison
from complete system (B). Reduced failures are due to better ventilation
is made of mean removals from
100 sockets during 100 hours of
operation when the tubes are re­
placed with new ones upon failure.

Electron Tubes The fourth column gives number

of failures if all no-defect removals
are eliminated; in column five,
eliminating both no-defect and elec­
trical-defect removals indicates un­
predictable f ail u res causing equip.
ment malfunction.
tenance pE'rsonnel. ThE' no-defect types constitute only about 10 per­ Information was also ob t ained
category includes tubes removed by cent of the types collected, they from commercial airlines on a num­
maintenance personrrel yet found represent nearly half the total re­ ber of 5654s made by diffe r ent
upon laboratory examination to con­ turns. Type 6AK5 alone contributed manufacturers.
tain no mechanical or electrical de­ almost 10 percent.
Type 6J6
fect s. Some tube types are numerous
As shown in Fig. 1, one third of among those returned due to the Tube t�·pe 6J6 is even more
all the tubes rejected from military large number in use. This group widely used than 6AK5 although
equipment are not defective at all. consists of old tYPfS such as the the num�er of sockets under
An additional one sixth of the tubes 6SN7, 6AC7, 6V6 and 6AG7 used in present surveillance is slightly less.
were probably damaged during well-established equipment or newer Table 11 shows the defect distribu­
installation or maintenance. The tube types having wide application tion of all returns of this type and
remaining 50 percent are truly de­ such as the 6AL5, 5R4GY and 6C4. i ts improved versions during the
fective. However, i of these failed A second category inciudes tubes first quarter of 1953. The no-defect
only after progressive deteriora­ whose large return is due mainly to category is larger for this type than
tion; they could have been located abnormal removals from some or for the overall average of all
and removed before causing trouble. all of the sockets in use. These in­ tuLes from all bases. In r eturns of
Therefore, only onE' out of every six clude types 6AR6, 6AT6, 6J4, the improved trpe, electrical defects
tubes represents a failure such as
would cause unpredictable equip­ SIX WAYS TO INCREASE ELECTRON-TUBE RELIABILITY
ment malfunction.
The equipment types installed at • Limit power dissipotion in soft-glass tubes to keep bulb temperalu;e below 200 C.
the different bases are similar.
• limit cathode temperature to values minimizing interface formation and sublimation
However, each base has a charac­
of nickel materiol.
teristic defect didribution pattern.
• limit heoter temperature to 1,400 K to prevent heater·cothode trouble and heater
These patterns reflect local operat­
burnout during cycling surges.
ing procedure and maintenance
practices . • Control metallic deposits on insulating surfaces by proper design of parts ond
insulators and proper shielding and coating of insulating parts.
Figure 2 is a ranking of the 20
tube types contributing the highest • Control dimensions to obtain proper fit of parts and reduce vibration.

number of returns during the first • Control fabricotion and ossembly to reduce operational failures due to faulty work­
quarter of 1953. While these tube manship

131 April, 1954 - ELECTRONICS

lead mechanical by two to one. A perature operation in its heat�r deterioration of this type, result­
breakdown of the electrical defects and cathode. This is indicated ing from evolution of gas and
is also given. by the large percentage of re­ poisoning of the emitting surface
A common electrical defect is 'turns for heater-cathode leakage by ion bombardment.
heater-to-cathode leakage, probably and insulation failures. The 12AT7 Failures may occur either due to
due to the unusual physical layout has had wide acceptance among loss of emission properties of the
of the tube. A large amount of circuit designers because of its high cathode, interface formation, evap­
power-as much as eight watts­ transconducta:1ce and low heater oration of materials. vibration or a

is dissipated between the cathode requirements in spite of the ex­ combination of any ·of these, and
and plates, and the grids have a treme design features used to ob­ by catastrophic chan�es such as

tendency to expand, usually toward tain such characteristics. These de­ gas discharge within the bulb, glass
the cathode. sign features are responsible for strains, and shorts and opens.
A comparison of results obtained the defects mentioned and should
in some controlled tests is given in
Deterioration Failures
outweigh the performance charac­
Table Ill. The figures are of the teristics if reliability is considered. The rate of d�cay of cathode
same order of magnitude as those The 2C51 is similar to the 12A'!'7, emission is related to the residual
for type 6AK5 and the improvement but except for some insulation fail­ gas in the envelope at the beginning
of the premium versions over the ures does not show the same weak­ of tube life. As time passes, the
prototype is also similar. However, nesses to as great an extent. The cathode absorbs some of this gas
the 6J6 appears to be overrated. 2C51 h as an improved version in and is gradually poisoned. This
To improve its reliability, it may the type 5670. process may be accelerated if addi­
be necessary either to decrease its tional gas is developed as a result
maximum rating or to redesign
Types 6V6 ancl 6AQ5 of electrical overloading or high
radically its mechanical structure. Tube type 6AQ5 has been ambient temperature.
A double triode such as the 6J6 studied in comparison with type Formation of interface rei!istal�ce
has almost twice as many chances 6V6 of which it is the miniature oc�urs most frequently in tubes that
of failure as a single tube but the counterpart. Table IV represents operate for long periods of time
returns do not show double the the defect distribution for these in circuits requiring low current
number of failures. This indicatei! two types and their reliable counter­ densities. Composition of the nichl
that double triodes, performing the parts. sleeve affects the speed of interfa�e
function of two tubes, have higher The 6V6 exhibits a larger per­ formation but exhaust schedules
reliability than two sin�le tubes. centage of electrical defects than of and operating temperatures are also
other defects and is much h igher important.
Double Triocles
in this category than the 6AQ5. The evaporation of metal from
Other double triodes. such as the The 6V6 shows a larger proportion hot elements in tubes produces
6SN7, 12AU7, 12AT7, 2C51 and of returns with interface resistance gradual deterioration in the in­
their improved counterparts. also while the 6AQ5 evidences a greater sulating properties of thl:' tll be
show higher reliability than com­ tendency to fail because of emission structure and low-resistan�e pat��
parable single tubes. degradation. between electrodes rClmlting in
The 12AU7 compares favor­ The temperaturl:' of the 6AQ5 has noise or irregular op�ration. Meth­
ably with the 6SN7 if the most been estimated to be 100 C above ods for minimizing thb effect are
recent design of the 12AU7 that of the 6V6 at maximum rating. coating smof'th surfaces wilh
is considered, Type 12AT7, on The higher operating temperature rough insulating matrrial to in­
the other hand, shows high-tem- may explain the faster emission crease the le � glh of the leakage

Table I-Removals of Tube Type GAK5 and Im­ Table II-Defect Distribution for Tube Type GIG and
pro',ed Versions from 100 Sockets During 100 Hours Improved Versions

95-perceDt 6J6 h.Jh\V.61)1)h hlol

Tube :\Iean confidence All Cal.as- Defect nllmlwr Iwro!nl 1It1l11lit'r l)l'rn'lI�

Sollfee type removals int�n'81 railures trophic

"lechAnic"lI .. . 107 �!l �o I"
failures
"
1(, 11
0'
Electrical .... . Hn '-

�arswdl "FII n,\ K5 :1.:16 2.06-5.00 1.7 0.90

t>09ti 1.73 0.89-2.8-1 0.n7 0.,10
No derect .. . �\1 10 .1.' n
.,
:\·Iiscdlnneou>; . 1:10 :I ,

:-;O[l :-;n,r"lk (a ir) r,AK5 .�. R, :I. 2,\.-�.�5 :I. (HI � 05

T.)I.,\' ..... , :':In 1110 ! �'t I!I/I
�fl5' 1,90 1.69-:1. H 0.71 0,17
n.\K.;\V \ .H:I \. 36-2.:lR 0.99 n. (Ih
. ----

r.omrn,·rcial airlinc�
J)f'�:ratlat ion. 1:1:; -) j
Tl,hl' mrgr A.li(lf� E 5tl:;l 0 �I 0.07-0. I:!
(',!!ulallr.1' . :17 l,i
Ttlt)t"lIlrgrA.lillc�n 5tl:;\ H.:I:! o 1�-O.nO
:�o
T"h" mfgr A . line I, 56H 0.17 0.05-0.:18 Ikal r-I·:.lllod,· It-aka�f·. 7(,

Tllla' mrKf B.linc D 565' 0.29 O. 10-0.60 T",,,I �IH loO

132 April, 1954 - ELECTRONICS

path, proper shielding of electrodes, Table nI-Removals of Tube Type 6J6 and Improved Versions from
elimination of getter material, use 100 Sockets During 100 Hours Operation
of n one vap or a tin g getters and in­
active alloys and limitation to
minimum operating temperatures 95-percent CaLas-
in all elements. Tube Mean confid enc e All trophic
type removals inter\'al failurQ!; failures
If tubes are operating under con­
Carswell AFB ...... . 6J6 '. . 7.0 5 .2 9. 1 2.8 '.67
ditions of constant vibration and
-

6099.6101 1.4 0.56 2 6 - . 0 .74 •. S&

frequent shocks, gradual deteriora­ NOB Nor folk (air) .... 6J6 4.46 3 4 5. 6
. - 3.22 2.12
NOB Norfolk (ships) .. 6J6 1. 86 1.3-2.5 1.39 0.67
tion can occur as a result of the
6101 2.2 1.0-3.8 1.55 0.66
enlargement of the sup p ortin g holes
in the mica spac ers. EventuaIly the
electrical output produced by
mechanical movement of the parts
Table IV-Defect Distribution for Tube Types 6AQ5, 6V6 and IDlproved
will overc o m e the useful output of
Versions
the tube. Corrective measures in­
clude an ch ori ng aIi elements firmly,
11sing close tolerances on the mica
6V6.6V6GT.
�pacers, processing the tubes at 6AQ5 6005,6095 6V6Y,6V6TY
low tempera tu re to prevpnt extreme Defect number percent num ber percent number percent
expansion, m;ing svnthetk mica, No defect ...... 152 49 4 17 45 41
Mechanical... ... 28 9 6 26 3 3
Terratex or ceramics for spacers
Electrical......... 79 2:> 4 17 51 47
enci re d e s i l!11 in g the structure to Miscellaneous..... 57 '7 9 40 10 9
ohtain shorter elertrodes and higher Total ........... 316 100 23 100 109 100
resonant frequencies for the ele­ Defective tubes only
ments. M echa ni ca l .. 28 17 6 32 3 5
Electrical ......... 79 49 'l 21 51 10
Some causes of destructive gas
Miscellaneous ..... 57 3·1 9 47 10 1:;
disch ar ge pressures are over heat­ Total. . ..... . ... . 164 100 19 100 6l 100
ing of Rn electrode until it re­
leases sufficient gas to reach arC
discharge pressures, hpa,\'Y heater­
cathode leakages and slow air leaks.
reliability in both type 6AS7G an d four-week period is readily notice ­

If the glass envelope is not an­

its reliable version, type 6080. Op­ able. Figure 3B shows return s of
nealed properly or is subj ected to
eration in hot climates with t h e all oth er tube t ypes in the complete
unusual mechanical or thermal
caLinet open gave sat i s facto ry re­ system d u ring the same period,
stresses it may crack However, sults if improved type 6080 wa s indicating that there was no com­
glass d efec ts are i n fr e q uen t where
used while opera t i on in a i r c on d i­
- paraLle change in the aggregate
tubes are carefuIly han dled. tioned vans gave normal life ex­ returns.
The electrode most subject to 'Pectancy for both improved and
shorts and opens is the heater. The Misapplications
JAN tube types .

operating tem pe rature is one of In the power supply uf a ra dar Severe mis a pplica tion s are read­
the major r e a s ons for failure. Low bombing system, the power supply ily detected and corrected and do
o pe r a t in g temperature for the not constitute serious reliability
chassis was originally mounted in
heater is important to avoid burn­ the nonpressurized part of the problems. More troublesome are
out a n d to reduce heater c ath ode -
plane where there w a s inadequate misapplications that produce only
leak a ge slight hardships. These include use
ventilation; 6AR6 return s per
.

O n ly about 12 pe rcen t of all tubes socket were four or five times of a tube type at its maximum rat­
c o l l ec ted had defini tely identifiable greater than the average for all ing when there is available an­
mechanical defects. Of these, more other sockets of the sy stem. During other type that could be operated at
than half were attri but able to the fiighh: ab ove 30,000 feet, bulb a more conservative r ating ; use of
heater or heater-cathode circuit. temperature reached 315 C. a type not actua lly designed for a
The power supply chassis wa s particular application when a m ore
Environmental Effects
transferred to the pressur ized zone suitable type is available; and use
High am bien t temperatures , vi­ of the plane a nd more effective of several tubes in parallel to avoid
bration and sh oc k and abnormal blowers installed. A temperat u re the use of a larger tube.
supply vo lt a ge contribute to un­ check in the relocated chassis
reliability. Ret urns of tube type showed an average bulb tempera ­
BIBLIOGRAPHY

6A.S7G. and its reliable version, the ture of181 C. Figu re 3A shows the "Investigation of Electron Tube Rell­
returns of tube types 6AR6 and
ab\1lty In Military Applications". ARCINC
6080, from a v o l tage regulator used Gen Rpt No. 1, GPO. Washington. D. .
6098 for four-week i n tervals over P. T. Weeks. Rellablllly In M lnla hsre
in a van mounted
- th e control and Submlnla.ture Tubes, Pr.o 11tH, 19.
!'�'stem illustrate this. a period of almost two years. The No. 6. p 499, May 1951.
G. H. MaUon. S. Wagf!ner, M. F. Holme.
Operation in hot climates wi th g ra du al decrease in re t urns from and M. R. Child. The Life ot Oxld ..
Cathodes In Modern Recelvtng Valve".
the ca bi net closed resulted in poor an average of 57.1 to 8.8 for each Proc lEE. 99. Part Ill. p 69. Milr�h 1952.

133 April, 1954 - ELECTRONICS

HI;AT-DISSIPATING ELECTRON TUBE SHIELDS

AND THEIR RELATION TO

TUBE LIFE AND EQUIPMENT RELIABILITY

JOHN C. MeADAM
Vice-President - Engineering

International Electronic Research Corporation

135 West Magnolia Boulevard

Burbank, California, U.S.A.

 HEAT-DISSIPATING ELECTRON TUBE 5HIELDS AND THEIR
RELATION TO TUBE LIFE AND EQUIPMENT RELIABILITY

The failure of electron tubes --- and the resulting failure of equipment --- is
a problem that has been long with us. So long, in fact, that many of us appear
to be condi·tioned to a state of mental resignation regarding it. High failure
rates have been accepted as a "fact of life", a condition which we find costly
and hazardous, but with which we must live in the electronic age.

Strangely --- for so general a condition --- our studies have shown a lack of
understanding of the nature of the problem. For example, while tube failure
rates are known to be high, the relationship between tube failures and equipment
failures is not so well known. Neither do we find broad understanding of the
reasons for tube failures.

Through this presentation we hope to show some significant facts:

1. That 3 eqUipment failures out of 4 are caused by

failure of tubes;
2. That the main cause of tube failure is high operating
temperature;
3. That an immediate solution to a major part of the
problem can be obtained through use of heat-dissipating
tube shields.

Toward the end of World War 11 and in the years since then, numerous surveys
have been made by the U. S. Military, either directly or through contracts
with research organizations, to determine the causes of electronic equipment
unreliability. These studies have shown that resistors cause about 3% of all
equipment failures, another 3% to 4% of failures can be traced to capacitors,
but the big cause of equipment failure --- and this is the unquestioned summary
conclusion of all the studies --- is failure of tubes. The reports show that
not. less than 75% of all equipment failures are caused by tube failures. (a)
And further investigations have shown that the main cause of tube failure is
high operating temperature.(b)

OUring the past five years, many additional tests have been conducted by
commercial and mil ttary laboratories. These tests have shown that the "normal"
operating temperatures of many tubes are well above safe limits.(c) In a
thumbnail analysis of these many surveys, the poin.ted conclusion is reached that

(a) - See Ref. 1 through 6 and 3 1

(b) - See Ref. 7 through 12
(c) - See Ref. 7, 8 and 10
 - 2 -

if a mere two-to-one increase in tube life could be achieved, it would eliminate

more than one-third of all equipment failures. Put another way, this much
improvement in tube life would accomplish more toward raising equipment
reliability than if all other causes of failure were completely eliminated.

Notable, also, is the effect that tube life improvement would have in reducing
equipment maintenance costs. During an average five-year use period, the cost
of maintaining electronic equipments, as calculated by U. S. Military services,
represents a minimum expenditure equal to 10 times the original cost of equip­
ment purchase.(d) (Fig. 1 ) That is, if an equipment is purchased for $10,000,
the maintenance cost for a five year period amounts to $100,000. And --- obviously
--- a saving of only 10% in the maintenance cost will save the entire original
purchase price,

JlllAlNTOW«;( COST
$ LOO,OOO • Y£AIIS

.90,000

• 80,000 .,.....

* 10,000

i 60,000 , 'EARS

$150,000

i 40,000 , 'EARS

t 30,000

$20,000 I YEAR

INITIAL Pt.IAQWII: C08T

• 10,000

MAINTENANCE COST OF MIUTARY EQUIPMENT

Figure 1

A U. S. Navy engineer reports(e) that his branch of the military service spends
$45 mill-ion annually for the purchase of replacement electron tubes. He also
indicates that it costs ten times this much to get the replacement tubes through
the channels of supply and into the equipment --- a total cost of $495 million
every year. The report calculates that if tube life could be prolonged by a
factor of only 2, some $247 million could be saved --- and if prolonged by a
factor of 4, then the yearly saving would amount to $371 million.

(d) - See Ref. 13, 14 and 30

(e) - See Ref. 15
 -3-

An illustration in terms of a single piece of equipment is reflected in a study

made by Aeronautical Radio, Inc.(f) This was a two year field test on equip-
ments using six type 6005 tubes. Under conditions of original application , the tubes
had an averate life of 1 000 hours. By doing nothing more than applying heat-dissipat­
ing tube shields to these tubes, average life was increased to 12,000 hours. With
six tubes on each piece of equipment, this means that no less than 66 tube failures
were eliminated for every one and one-half years (12,000 hours) of operation. At
an average cost of $5.50, the purchase saving on the tubes during such a one and
one-half year period would amount to $363 --- and this saving is accomplished by
an expenditure of approximately $9 for the six "Life Saving" tube shields.

I hope I will be forgiven for using the term "life saving". While I was applying
it to tube life, I know all of us are aware of the human-life-saving aspects of
electroniC reliability - whether it is from the certain functioning of a warning
system, the accurate firing of a defensive weapon, or the correct radar intelligence
that may save the life of a single aircraft pilot. A better appreciation of what
1s involved may come from deeper penetration into:

THE SCOPE OF THE PROBLEM

Our present predicament with electronic equipments, and with the tubes they use,
has grown with our technological development of complex electronic systems. Before
World War 11, a typical destroyer carried equipment using about 60 tubes; a modern
destroyer requires 4, 000 tubes for the operation of essential equipment. As
requirements for equipment have gon€ up, space available for individual equipments
has almost disappeared. Equipment has been miniaturized and then subminiaturized .
And the smaller equipment is asked to perform more functions than the old .

Each change has meant more heat generated with less space in which it could be
dissipated. The smaller tubes often generate as much heat, individually, as
their older counterparts; yet they are crowded into ever smaller spaces, bundled
together , causing an ever increasing environmental temperature in which they must
operate.

The degree of temperature increases to which I refer is reflected in tests

conducted by the U. S. Navy Electronics Laboratory. These tests show that the
internal temperature rise in such conventional equipments as radar, communication
and navigation, etc . , amounts to 30 to 40°C. This must be added to the required
o
external environment of 50 to 8S C for military equipment and means that internal
environmental temperatures run at about 1000C for many of these equipments.(g)
Curves, to be illustrated later, show that some tubes exceed their maximum bulb
temperature ratings when operated in an ambient temperature of only 2SoC.

These discrepancies in ratings versus actual operating conditions bring to light

a serious aspect of the total problem - the lack of adequate temperature-li.fe
information on tubes. This data should be available to design engineers. Yet,
neither specifications, nor manufacturers' data sheets, list pertinent te�perature
data on tubes.

(f) - See Ref. 1 6

(g) - See Ref. 1 9
 �-

For some tubes, a maximum allowable bulb temperature is listed by the manufacturer.
But this, frankly, is a poor guide when reliability is at stake. Usually it
represents a maximum temperature at which the tube will pass certain specification
tests . .. �ut if this same temperature is prolonged - as in actual operation - it may
well indhce rapid failure. It will be shown later that the temperatures listed in
many specifications are not only vague and misleading, but actually promote the
misapplication of many tubes.

Another negative factor that has evolved with increased use of electronic equipment
is the widespread use of what is referred to in the United States as a JAN-type
shield. (Figure 2) This shield, developed a number of years ago to afford
electrostatic shielding and retention for miniature tubes, is probably responsible
for more tube failures than any other single cause.

Figure 2

Figure 3 shows that bulb temperatures, when the JAN shield is applied, range
more than lOOoC higher than bare bulb temperatures operating in the same environ­
ment. The solid wall of the JAN shield, the captive air between tube and shield
and the shiny, heat-reflective finish, mean that JAN shields increase the
temperature of all tubes with which they are used - without exception.
 -5-

420

U
•

Z
w
a:
::>
t:
a:
w
Cl.
�
w
I-
(0
--'
::>
(0

'
4'0 60 80 100 120 140 160 IS O

AMBIENT TEMPERATURE IN ·C
GOD5/6AQ5 OPERATING AT MAXIMUM PLATE DISSIPATION .

F i gure 3

To summarize the scope of our problem, we are dealing with more heat than ever
before, 1n smaller spaces, in vastly increased numbers of equipment, and in
infinitely more cri t i cal applications. Compounding these factors, we have a
lack of temperature-versus - life data on specific tubes and a general usage of
a type of electrostat ic shield which rai ses tube bulh temperatures from 20 to
35% above normal.

STUDIES LEADING TOWARD SOLUTION

There has been general agreement for some years that high operating temperatures
cause tube failure. It i s more recent that h i gh temperature has been identified
as the main cause. And, unfortunately, there has been very little precise test
data to indicate the relationship of tube life to bulb temperature.

In 1 954, the U. S . Army Signal Corps Laboratory gave a contract to the General
Electric Company at Owensboro , Kentucky for a life data study on electron tubes.(h)
One factor tested was the influence of bulb temperatures. Five types of min iature
tubes were selected and lots of 200 tubes of each type were tested. The lots
consi sted of tubes from several different manufacturers. Each lot of tubes was
operated at a selected bulb tempe�ature for a period of 5000 hours. The per-
cent surviving at various time intervals within th i s period was plotted in
relation to the operating temperatures .

F i gure 4 shows the failure rates of Type 6005 tubes at five pre-selected constant
operating temperature s . Note that while 97% of the tubes tested at 2200C were
still operating at the end of the 5000 hour period, an increase of just 1 70C --­

to 2370C --- produces a failure rate of 32% in the first 2000 hours. Also not,e
that on the lot operated at 3l60C , 98% failed in the first 2000 hours. Yet 3160C
i s lower than the temperature at which the same tube operates in a JAN shield at
room ambient.

(h) - See Ref. 9

 -6-

200 TUBE LOTS 6005/6AQ5W

OPERATING TIME IN HOURS

Figure 4

Figure 5 shows similar data on the 6AK5 tube. Here again the pattern is the
same --- failure rates always increase as bulb temperatures rise. Other curves
from the General Electric Report show the same pattern, although time and space
do not permit me to show them.

200 TUBE LOTS 5654/6AK5W

10 o -

i I
""'"f-. 1/
I

" i

"-
0

-\; f::: f't......

V1'\
.....
(}

� f-,..
I

1'\
0 92

;--.f-.
.
I'"
-
t- 312
0

0
1000 2000 3000 4000 !5000

OPERATING TIME IN HOURS

Figure 5
 -7-

One of the other tubes covered by the report was a Type 5670 which operates at
a low dissipation of less than one watt. This tube showed a definite failure
increase after 3000 hours even when operated at the low temperature of 1150C.
Specification and tube data sheets list 1650C as the maximum allowable bulb
temperature for this tube. This fact casts light on a related problem of mis­
application in equipments which contain many tubes. When the tubes are of
different types, although some may withstand reasonably high temperatures, others
operating nearby may suffer sharply curtailed life. It will be shown later
that the most practical approach to the solution of this problem is the use of
the most efficient heat-reducing device on all the tubes, regardless of their
individual temperature limitations.

Further studies have been under way in recent years to determine the effect of
temperature on the materials and characteristics of electron tubes. Cornell
University Engineering Laboratories ( i ) , Sylvania Electric Company ( j) , and
Aeronautical Radio, Inc. (k) , all under military contract, were prominent 1 n
these investigations. A summary o f their findings follows:

Deterioration of tube performance characteristics as the

result of elevated temperatures is principally caused by
evolution of gas within the tube. Other deterioration
causes induced by high temperature, in order of importance,
are: getter migration, grid emission, glass failure, inter­
electrode leakage, contamination , grid loading, and loss of
emission.

The gas evolution is from the inner surface of the glass and
the surfaces of the plate and other tube elements. Its
occurrence varies directly with temperature increase. The
gas "poisons" the tube and causes a gradual lowering of its
transconductance --- to the point of ultimate failure.

°
CATHODE -SOO C

°
GLASS BULB-115 C
GRIO- 500·C

'
PLATE-400 C

TYPICAL MINIATURE TUBE TEMPERATURES

Figure 6

( i ) - See Ref. 10 and 20

(j) - See Ref. 21
( k ) - See Ref. 8
 - 8 -

High temperatures are necessary on the f ilament and cathode elements of the
tube as shown i n Figure 6. There i s no necessity for heat to be present on
the other elements. In many tubes the f i lament d i s sipation i s only a small
part of the total d i s s i pation of the tube. A 6AQ5 tube , for example, has a
plate d i s s i pation of about 1 2 watts, but only 3 watts of f i lament d i s sipation.
It is the combination of the heat generated on the plate of the tube --- and
the ambient temperature --- which causes the tube glass and other elements in
the tube to operate at exces s i ve temperatures . The type of glass used i n most
m i n iature tubes is such that i t absorbs most of the heat radiated by the plate.
A high percentage of the tube beat is concentrated on a very small area of the
glass, as shown in Figure 7 .

2.00 J�
�
�
1.75
0

1.50

j" .. ...
-
0

R�TATION SURVEY
fJ) 1.25 0

W '\ 1 0

'.'
___ 900
J: ,.oO
U
0

Z 0

0.75 ,... 0
..,- 0
- -- �...
�
0.50
...... 0
'"
...
�
0.25
0

o 140 160 180 200 220 240 260

SURFACE TEMPERATURE IN C·
6005/6AQ5 OPERATING AT MAXIMUM PLATE DISSIPATION
(45' AM81ENT TEMPERATURE)

:Figure 7

This "hot spot" appears opposite the center of the plate.(m) As the ambient
temperature increases, bulb and plate temperatures are increased in turn, as
well as the temperatures of other tube elements. Thi s great d ifference i n
temperature between the relatively cool ends o f the tube and the hot spot at
its center is the main cause of glass failure.

(m) - See Ref . 18

 - 9-

I��
lO_ "'"
,---

i '\.. ..

....""'-
!- . .
..

e-r "-
.
TEMP. WITH OAR<
FORC!D AIR
\
.J
BWI

r-- COOL.JNG moP. .

.... ..-- .. .
r--
...

I.
�.
BULB TEMPERATURE UIlI UI
GRAOIENT ON TUBE SURFACE

TUBE SURFACE TEMPERATURE GRADIENT WITH FORCED AIR COOLING

F igure 8

Most past attempts to cure the heat problem have involved forced air convection.
As shown i n F i gure 8, forced a i r cooling lowers the ambient environment of the
tube - providing a partial solution - but does little to relieve the temperature
gradient present on the tube surface. Also, because of mechanical difficulties
in d i recting air flows, tubes are often found in air pockets, Figure 9, so they
receive little cooling effect.

INEFFICIENT FORCED AIR COOLING ....

Figure 9
 - 10-

WHAT ABOUT TRANSISTORS?

Some, who have given up any hope of solving the tube heat failure problem, are
patiently waiting for a day when transistors and magnetic amplifiers may replace
all tubes. There is little doubt that eventual development and production of
such components will solve many of our present problems. But transistors have
heat problems of their own; they appear to need many additional years of
development and production experience before they can hope to replace tubes in
their hundreds of millions of present applications. Since we must live with
tubes for years to come, then we must live with the problem of dissipating the
heat. An effective solution to the heat problem has actually been available
for the past five years - though it has not been widely known.

SOLUTION: THE HEAT-DISS IPATING TUBE SHIELD

About five years ago, International Electronic Research Corporation ( IERC)

developed an effective heat-dissipating tube shield , Figure 1 0

Figure 1 0

This tube shield dissipates the heat by radiation, convection and conduction.
It grasps the hot tube bulb and distributes the heat from the hot spot over a
large surface area. This way , it not only reduces the general temperatures
present on the tube glass , but also greatly reduces the temperature gradient
along the surface of the tube, Figure 11
 -11-

LOW MIGH

'-
, -�
-- ,I-
I
9
"

�
•

!
�
BARE BULB TEMP.

I'
•
.. ..
-
•
>
..
�I. .
• �
BULB TEMPERATURE

GRADIENT ON TUBE SURFACE

TUBE SURFACE TEMPERATURE GRADIENT

WITH HEAT-DISSIPATING SHIELO COOLING

:Figure 11"

As shown in F i gure 3 , the temperature reduction when using and I.E.R.C. Type B
0
shield on a typical 6005 tube amounts to from 40 to 50 C below the bare bulb
0
operating temperatures and from 125 to 150 C below corresponding operat ing
temperatures encountered with the use of the old JAN shi eld.

It should be remembered that the JAN shield came i nto wide use before the heat
failure problem of electron tubes was generally recognized. But it must be
observed that i f one were purposely to design a poor thermal device for a tube
shield - a literal "k iller shield" - he could do no better than the JAN shield
as i t was originally designed and is still used. As earlier mentioned, the JAN
shield assures a capt ive a i r space between tube and shield; its shiny surface
reflects heat back into the tube. In addition , there is almost no thermal contact
from the shield to base. or from the base to the chassis.

Design of the IERC heat-dissipating shi elds began from the base up - in recognition
of all the defici encies i n the JAN type shi elds. In the M i litary Type B shi eld
( F igure 10) , for example, the base has a wide flange for effective thermal
contact ..,i th the chassis. The shield, in turn, has a tight thermal f i t to the base.
The tube i s grasped firmly along its entire surface ..,ith the beryllium copper
spring f i ngers. To date , this type of shield has proved to be the most effect ive
vertical mount type shield - both i n its heat dissipation characteristics and its
retention of the tube i n extreme shock and vibration �nvironments. (n)

About th."ee years ago, need arose for a method of applying heat dissipating shields
in equipments which had the old JAN bases mounted to their chassis. It did not
appear practical to remove sockets from these thousands of equipments i n order to
apply the M i litary Type B base and shield.

(n) - See Ref. 22 and 23

 -12-

Figure 12

Figure 12 shows the retrofit type shield which was developed by I . E.R.C. to meet
this problem. This shield, with its beryllium copper spring finger liner and its
blackened shell, snaps onto the old JAN type base. Although it is a little less
effective than the Type B shield - primarily because of lack of thermal contact
between the JAN base and the chassis - its performance characteristics as shown
in Figure 3 indicate that it does reduce temperature of the tube bulb well below
bare bulb temperature - and more than lOO oC below JAN shield temperatures.

While it has been shown ,that high temperature shortens tube life hazardously -
and that the heat dissipating tube shields as described effectively lower tube
temperatures - correlation of these facts is needed to prove that heat dissipating
tube shields prolong tube life and increase equipment reliability to a high degree.
 -13-

100 IERe (TYPE B

...
UJeo
'" ;;;;-
;;;
;! 'ERe (TRJ.
z
UJ
U60 -
a:
UJ
"'-
\ -
-' 4
0
�
....r-...
..
;; ......
a:

i;l,0 �
,
,,� ,"I"n I
BARE BUL
B

o
500 1000 1500 2000 2500

OPERATING TIME IN HOURS

6005/6AQ5W - TOTAL DIS SIPATION
16.4 WATT S IN 100· AMBIENT B ARE BULB TEMPERATURE 270"C

Figure 13

Figure 13 compares the failure rates of 6005 type tubes when different types of
shields are used, This information is derived from the data shown in Figure 3
(Tube temperatures vs. tube shields) and the information contained in the curves
in Figure 4 (Tube life vs, tube temperatures),

100
I
i r--,-

.......
+-
r--.. t-;....
--
..... .....
0
r
Ee

I- I
1'-....
R

t-
(TRJ

j.."
0 r-...
.......
r-..
!--...
I

�i--
0

0 1-- SOD
- I

1000 1500
! i
2000 2500
OPERATING TIME IN HOUlIS
5654/6AI<5W - TOTAL DISSIPATION
2,97 WATTS IN 100·C AMBE
I NT BARE BULB TEMPERATURE 190·C

Figure 14

Figure 14 .wows this same correl ation for a 6AK5W type tube, The extended life
Rftd reault1ng increase in rel iability is clearly indicated,
 -14-

There have been other attempts to solve the problem of heat dissipation.

Figure 15

Figure 15 permits some comparisons. At the left is a cut-away of a JAN shield.

Next to it is a JAN Shield cut-away showing a corrugated brass insert liner.
This corrugated liner represents a different attempt to relieve the tube
temperature problems caused by the JAN shield - the liner being intended to
conduct heat from the glass to the shield. Although some laboratory tests indicated
this liner was an effective device, (o) quantity use in field service showed it to
be of doubtful value. Variations in diameter of the glass bulb in tubes of like
identification caused the trouble. Diameters of seven pin miniature tubes were
found to vary from .700" to .730" and variations were from .SOO" to .S40" in nine
pin tubes. For the liner to be effective it must contact the tube bulb surface -
and hence must conform to the diameter variations. But the brass corrugated type
liner- because of its configuration - cannot act as a spring.

Two deficiencies have resulted: when placed over tubes of larger diameter, either
the tube breaks or the liner is permanently deformed; when a deformed liner is
placed over a tube of smaller diameter - a frequent occurence in field use when
tubes are replaced and shields are re-used - the heat conductance from tube to.
shield is no longer effective. Of interest is the fact that the U. S. Navy
Department, which originally promoted use of this brass corrugated liner , no longer
recommends its use. It has become obvious that tube shields for heat dissipation
must have liners with spring properties.

OTHER REQUIREMENTS OF TUBE SH IELDS

To this pOint, I have discussed the heat-dissipating characteristics required of -

and supplied by - I.E.R.C. shields. A shield must also perform in terms of
electrostatic shielding and mechanical retention of the tube. The solid cover

( 0) - See Ref. IS
 ------ ------

-15-

provided by the JAN shield is not essential to proper e l ectrostatic shielding.

This probably arose for no other reason than convenience of manufacture. A
screen or open-work shie l d of solid metal, if well grounded, stops el ectrostatic
radiations. All of the I.E.R . C. shields have proved el ectrostatically effective .

The problem of tube retention is a little more complex. In applications such

as those encountered on aircraft and missiles, it is necessary that the shield
retain the tube in environments of extreme shock and vibration . The Mil itary
Type 8 (p) and the TR Type(r) have been tested to 2000 cycles of vibration and 90
G's of shock. 2000 cyc les is the l imit of present tube shie l d specification
requirements. (5) 90 G's of shock was the limit of the test machine. There are
no U. S. Military specifications for shock tests on tube shields . There is no
doubt that these shields will exceed these limits.

Studies by Aeronautical Radio, Inc.(t) rate vibration second only to high

temperature as a cause of tube failure. It is interesting to relate the effect
o f the type of tube shields being discussed to this vibration problem .

A test was conducted to determine the dampening ef fect of the spring finger liners
in the shie1ds. (u) This test was made in accordance with the micro-phonics test
described in MIL-E-l. A signal was fed into a tube which was under vibration;
then the variation in the amplitude of tube output was measured. Results showed
a 30% reduction in the variation when the tube was retained by a shield with a
spring finger liner, compared with an unshielded or JAN-shielded tube.

CHECKING TUBE SHIELD EFFECTIVENESS

So far, our discussion has concerned itsel f with the need for heat dissipation
to reduce tube and equipment failures and with the qualities and characteristics
of the heat dissipation tube shields designed to meet the problem. Important for
the user are uniform methods of test by which product standards may be established
and product performance verified .

The Cornell Aeronautical Laboratory(v) has been investigating tube temperature

and tube shie l d problems for the past several years. The laboratory - by measuring
tube el ement temperatures within the tube - found instances in which certain
shields , intended for heat dissipation, were apparentl y effective for reducing
bulb surface temperature, but were actuall y reflecting heat back through the
gl ass with resulting increases in the plate temperature. The use of standard
tubes , modified for internal temperature measurement, was found to be a tricky
and dif ficult procedure , with added problems caused by variations in individual
tubes. So , in search of more consistent results and to facil itate further
testing , the Corne l l University E l ectrical Engineering Laboratories developed
instrumented tubes which they call "Thermatrons". (w) (Figure 16)

- - - - - - - - -
- -

(p) -
See Ref. 22
(r) - See Ref . 25
(s) - See Re f . 26
( t) - See Ref. 8
(u) - See Ref . 25
(v) - See Ref. 20
(w) - See Ref. 27
 -16-

HIGH WATTAGE HEATER

PLATE THERMOCOUPLE

BULB THERMOCOUPLE

CORNELL THERMATRON

Figure 16

These Thermatrons are standard tubes with the filament and cathode elements
removed and replaced with a heavy duty heater which can dissipate heat equal
to the total amount of heat originally dissipated by the filament, plate and
other elements. The laboratory found good correlation between the internal
operating temperatures of the tube, using this method, as compared to st·andard
tube operation . Thermacouples, attached to the plate and the inside of the
glass bulb, are brought out through the unused pin leads. Using this method,
the wattage of the tube can be easily varied and the plate and bulb temperatures
can be monitored during tube shield tests.

Carnell Aeronautical Laboratory has also determined that - since heat conduction
was of prime importance in evaluating tube shields which are mounted to chassis
and heat sinks - correct data cannot be obtained unless an infinite or constant
temperature heat sink is used during the test. They devised an infinite heat
sink, shown in Figure 17.
 -17-

CONSTANT TEMPER ATURE HEAT SINK

Figure 17

This consists of a heavy copper chassis standing in circulating water and enclosed
in a clear plastic case to eliminate uncontrolled convection currents. The use of
these methods should eliminate much controversy and confusion in testing tube
temperatures and tube shields.

At the present time , U. S. Military specifications ( x) require that the heat

dissipating properties of tube shields be determined in the following manner:

An aluminum slug the shape of a tube and containing a heating element in its
center and thermocouples embedded in its surface is raised to and stabilized at
a temperature of l800C. The appropriate shield is then placed on the slug and
the temperature reduction of the slug is noted. This slug is mounted in a tube
socket in the center of an aluminum sheet metal chassis four (4) inches square
and two ( 2 ) inches high with the appropriate base also mounted with the socket.
This method was devised about five years ago for the want of a better standard
method and was placed in the specification .

It seems apparent to many that a more sUitable test method is needed. The method
outlined above has the following specific faults:

1) The metal test slug is made of heat conducting mat,erial and assumes an even
temperature over its entire surface. The glass surface of an actual tube , by
contrast, is non-conducting and has at its center a high temperature spot as
described earlier in this paper. Obviously, a shield designed to lower the
temperature of the test slug would not necessarily meet the problem of the hot
spot temperature on a glass bulb .
2 ) Since , as has been noted , certain shields may lower the glass temperature of
a tube, but reflect heat back through the glass to increase plate temperature, it
is also obvious that the metal slug test provides no means for ascertaining this
fault.

(x) - See Ref. 26

 -18-

3) The specification method provides for use of a small test chassis, four inches
square and two inches high. Since one of the heat dissipating properties of a
well designed tube shield is conduction - the heat being conducted from tube
to shield to base to chassis or heat sink - the small chassiS is deficient in
its ability to reflect accurate comparisons of conductive properties . This fact
appears readily when it is realized that the amount of heat conducted between
materials 1n thermal contact with each other depends upon the temperature
difference. When a small chassis is used, the efficient heat conducting shield
will raise the temperature of the small chassis substantially more than will the
less efficient shield . Thus, the less efficient shield gains the benefit of a
greater temperature difference between shield and chassis. These circumstances
distort performance curves on both the efficient and inefficient shields so that
effective comparisons are not possible. The true values of various shields cannot
be tested without a test chassis or heat sink that simulates a standard condition .

A further note of warning needs to be stated with reference to evaluation of

chassis temperature determinations in connection with heat dissipating shields .
In speaking of the conductance of heat to the chassis by an efficient heat
dissipating tube shield, the impression may be created that effective shields
will so heat up the chassis as to risk damage to other components . While this
might be dismissed in view of our earlier determination that only a small
fraction of all equipment failures are traceable to failure of other components,
it seems more to the point to demonstrate that the conclusion is untrue .

Higher chassis temperatures actually result with the use of non-conductive JAN
shields and bases. The reason is that the JAN shield increases the temperature
of the tube by many degrees . There is then a higher hot spot temperature
available in close proximity to the chassis. This condition can, and does. cause
a higher chassis temperature around the tube.(y) A comparison of this temperature
condition with that of the Type 8 heat dissipating shield is shown in Figure 18.

IOO·C

JAN SHIELD TYPE B SHIELD

CHASSIS TEMPERATURE- JAN SHIELD VERSUS HEAT-DISSIPATING SHIELD

Figure 18:

(y) - See Ref . 29

 -19-

The Type B Shield , as shown, produces a lower chassis temperature in the vicinity
of the tube. This is because it lowers the hot spot temperature of the tube
and spreads the heat over the surface of the shield from which it is dissipated
by radiation and convection as well as by the conductance through the base to
the chassis.

FURTHER DEVELOPMENTS IN HEAT DISSIPATION

The heat dissipating tube shields discussed so far are vertically mounted shields
for use in conventionally constructed equipment.

Figure 19

As you will note (Figure 19), there are thermal problems scattered throughout
most such equipment . Many equipments will continue to be constructed in the
future as they are today. The heat dissipating shields we have discussed are
available now to eliminate a significant percentage of the unreliability problem
in these equipments.

Such shields as we have discussed, however, cannot be considered the end of the
matter. Tubes are the most unreliable component in an equipment and they also
represent the majority of high temperature parts. This leads us to recognize
that tubes need to be given prime consideration, not alone for what can be done
to lower their temperatures after an equipment has been deSigned, but in the
initial thermal design of the eqUipment, itself .

I. E.R.e. has been engaged for the past five years in a continuous program of
research and development to improve electron tube and equipment thermal problems.
We feel this work has been productive:
 -20-

CD SHELL 8 BASE

<ID TUBE IN SHE LL

@ SPRING CLIP OVER SHELL 8 TUBE

HORIZONTAL TYPE WRAP-AROUND SHIELD • • • • •

Figure 20

The heat dissipating tube shield shown in Figure 20 mounts horizontally to a heat
sink. This shield, with its soft wrap-around shell in intimate contact with the
tube glass, conducts the heat to its base.

SOF T WRAP-AROUND SHELL

C ONF ORMING TO IRREGULAR TUBE SURFACE

Figure 21

In Figure 21 , you will note that in addition to the effective heat dissipating
properties of this shield, it provides an excellent retainer of the tube in
 -21-

extreme shock and vibration environments as it pl aces the tube in a position with
a low center of gravity . And because the soft wrap-around shell conforms to the
glass and forms a cushion between the hard glass and the hard spring clip there
are no resonances set up in vibration.

This assembly, with its base attached to a heat sink as shown, wil l maintain the
tube bulb at less than 50C above the heat sink temperature per watt of dissipation
in the tube, This is shown in Figure 22.

300 .---�-----r--'---'---��-'

o
U250�----�-----+--��--+---1
z

�2001----1--�---4--�-+----����--4---���--
Cfl

1 0�----�-----+��-+����--f---t-1
I­
�
W

5
I

�IOO�--�----��-����-
W

CD
�
W

�Cl�ME
Cfla::

5
35 SH1E 0
WITH I Rc 1 5-

°
(} 2 4 6 8 10 I I 16
6005/6AQ5 TOTAL DISSIPA TION IN WA T TS

Figure 22

When al l the tubes are mounted to a heat sink as shown in Figure 23, this
eliminates the need for scattering heat generating tubes throughout the equipment;
it isolates the problem by transferring the heat into the heat sink_ These
conditions greatly simplify the thermal problem_ Al l the heat is concentrated
in one area and can be transferred in simpl e manner from the equipment by various
methods, dependent on the volume of heat present.
 -22-

Figure 23

If the amount of heat to be dissipated is in a low range, the heat sink may be
a simple metal plate of sufficient thickness to conduct the heat away to the
frame of the equipment or to some other point as shown in Figure 24 .

CD

TYPES OF HEAT SINKS

J- CONDUCTION Z- AIR COOLING 3--L..- COOL"'"

Figure 24
 -23-

If the amount of heat to be dissipated is in the medium range, the heat sink may
be a hollow duct, possibly containing fins, through which cooling air can be
passed to transfer the heat. (This is an efficient method of using forced air
for cooling, contrasted with blowing air, at random, inside an equipment.) If the
amount of heat to be dissipated is in the high range, the heat sink may contain
ducts for the circulation of liquid coolants.

This type of arrangement solves two problems: it separates the electron tubes -
which are the main source of heat in an equipment - so that thermal problems of
other components are greatly reduced; it also places the tubes in a condition
where their temperature can be maintained most effectively at as Iow a point
as desirable.

All of the shields shown so far have been for application to miniature and sub­
miniature tubes. The same problem of heat dissipation and mechanical retention
exists with tubes of the larger octal and Dower tube sizes.

Figure 25

Figure 25 shows a shiel d which has a Beryllium copper spring finger liner for
grasping the tube and transferring the heat to the shell. The shell fastens to
the base with a positive snap action lock . Note that the base has a good solid
thermal and mechanical contact to the chassis or heat sink. This shield will
retain the tube at 2000 cycles of vibration and 90 G's of shock, as well as
meeting all of the general tube shield requirements of heat dissipation, salt
spray, and heat resistance, installation and withdrawal forces, etc. of the
Military Tube Shield Specification M IL-S-9372 (z) .

ARINC has been testing the Tl2 shield shown in Figure 25 on 6080 type tubes
in radio receivers and on 5932 type tubes in radar repeaters. They report in
a letter dated September 12, 1957 that the use of these T12 shields on the

( z) - See Ref. 31
 -24-

eight 5932 tubes in each radar repeater decreases the failure rate of these tubes
from 36% to 12% in a 10,000 hour period. In other words, the 24 failures that
had been occurring in each equipment every 10, 000 hours were reduced to 8 failures
during the same period when the heat dissipating shields were applied.

The 5932 tube costs approximately $10.00 each, so that $120.00 worth of tubes
are saved each 10,000 hours, or about $600.00 worth in a five year period for
each radar repeater. This saving considers only the tube cost and not the
associated maintenance saving, nor the elimination of down time of the equipment
in time of need.

CONCLUSION

Many persons concerned with electronic reliability have perhaps jumped to a

conclusion that since tube shields are relatively simple devices they need little
consideration. Among many there has been a tendency to concentrate on the
inherent properties of tubes --- even though millions of dollars in research in
this direction have produced relatively minor increases in efficiency and
reliability.

We hope that this presentation has served to show that the problem of heat­
dissipating tube shields involves a variety of considerations: the effects on
internal temperature, the types of mount, the capability of retention in shock
and vibration environments, the methods of test, the methods of disposing of
heat from the area of the equipment, and many other points. We would be remiss
if we were regarded as indicating ultimate answers have been reached in any
direction. Yet, in spite of the confusions which still exist, we feel we are
accurate in stating that the heat dissipating tube shields discussed do represent
an important solution to a major part of the electronic reliability problem.

In the United States the confusion is gradually being resolved. The U . S. Air
Force in 1953 tested the Military Type B shield (aa) after conducting other tests
which showed the alarming temperature increase caused by the JAN type shield. As
a result of these tests they wrote a heat dissipating tube shield speCification,
M IL-S-9372 (USAF) in 1954. This described the Military Type B shield. Until
recently , use of the JAN shields was still permitted. In January 1958, the
Air Force cancelled future use of JAN type shields, and established requirements
that M IL-S-9372 shields be used. The shields covered by this specification have
been illustrated. Figure 10 shows the Type B shield now specified for new
equipment and Figure 12 shows the Type TR which is specified for retrofitting.

But this specification still does not describe shields for tubes other than
vertically mounted miniature tubes. No consideration is given subminiature,
horizontal miniature, or octal and power type shields.

An attempt was made about two years ago to coordinate this specification among
the three U. S. Military services, the Navy, Army and Air Force. But because
of personnel turnover and resulting confusion respecting data on the subject,
no conclusion was reached. The Navy, pressed by its own need for heat dissipating

(aa) - See Ref. 28

 -25-

tube shields, then undertook its own specifications, MIL-S-19786(Navy). This

specification is less comprehensive than that of the Air Force.

During the past five years equipment manufacturers who recognized the problem
and wer� willing to put forth an effort on their own to bring about a solution
in spite of the confused specification reqUirements, have readily accepted and
applied these shields in their equipment. It is felt that a good degree of
the success of many missile projects are due to the application of the shields
described here. Designers of such missiles as the Matador, Regulus, Talos,
Titan, Nike, Atlas, Sparrow, Hawk, Somare and many others have solved their
electron tube thermal and mechanical mounting problems through the use of these
shields.

In addition, many missile ground control and telemetering system designers have
recognized and applied this significant reliability aid. Other ground and air­
borne equipment designers have likewise recognized the value and importance of
applying these heat dissipating shields. It is significant that they have done
so without military reqUirement, but because of their interest and deep concern
over performance of their equipment.

Still other companies - aware of the heat problem, but unaware of the sheilds
on the market which could solve it - have sought solutions of their own. Many
of these attempts have lacked success for the reason indicated earlier - lack of
full understanding as to the complexities involved.

We hope we have helped to clarify some of these complexities and to provide data
from which you may reach more useful conclUsions than in the past. To move forwar.
we shall need effective standardization - in defining the job that heat dissip3tin
shields must accomplish, in setting manufacturing specifications, sizes and test
data. Meanwhile, with so much at stake in the reliability of so much vital
electronic equipment, we trust that a deep and concerned interest will be taken
by all involved.
 REFERENCES

1. Given F. J. "ReI iabil1 ty in Electronic Equipment", Guided Missile Committee

of AlA, White Sands Proving Ground, April 20, 1954

2. "Progress Report on Reliability of Electronic Equipment" Volume 1 8< 2,

Research and Development Board, Department of Defense, EL/217, Feb.ruary 1952

3. "Parts Failure Analysis", Vitro Corporation of America, Silver Springs,

Md., Report No. 25, November 1, 1951

4. Carbart R.R. "Survey of Current Status of Electronic Reliability Program",

Report No. RM-1131, The Rand Corporation, Santa Monica, California,
Augus t 14, 1953

5. NEL Reliability Design Handbook, Section 1.1, "Reliability Surveys",

May 1, 1955, U.S. Navy Electronics Laboratory, San Diego, California

6. "Reliability Factors for Ground Electronic Equipment" McGraw-Hi11 Book Co.,

for Rome Air Development Center, Griffiss Air Force Base, New York

7. NEL ReI iability Design Handbook, Section I, Page 11, "Electronic Equipment
Failure During Evaluation Test", R.J. Steelman, August 1, 1955

8. Aeronautical RadiO, Inc., General Report No. 1, "Investigation of Electron

Tube ReI iability in Mili tary Applications" Washington D.C., January 4, 1954
I,' I
I ' I
9. Bowie W.S. "Study of Electron Tube Life", General Electric Co., Owensboro,
Kentucky Contract DA-36-039 -SC-42524

10. Corne11 University, "Signal Corps Tube Analysis Program" Technical Report #22,
Cornell School of Electrical Engineering, Ithaca, New York, May I, 1955

11. Acheson, Marcus A. and Eleanor M. McElwee, "Concerning the Reliabil ity of
Electron Tubes", Proceedings of the IRE, Volume 40, No. 10, October 1952

12. "Electron Tube Temperature Problem in Relation to Design for Reliability",

Vitro Corporation of America, Silver Springs, Md., Report No. 51, November 30,
1953 (Contract DA-36-034-0RD-k426)

13. The Background of "Reliability", T A. Smith, IRE Transactions on Reliability

and Quality Control, September 1956.

14. NEL Reliability Design Handbook, "Introduction" by R.J. Steelman, March 4, 1954

15. "Heat Dissipation in Electron Tubes", W.D. Campbe11, Proceedings of the RETMA
Symposium on Applied Reliability, University of Southern California,
December 19-20, 1956

16. Aeronautical Radio, Inc., Report No. DCA 56-1161 "The Cooling Effects of a
Special Tube Shield", Field Engineering Report, Army Radio Station "WAR",
October 25, 1956

17. "Techniques for Reliability Measurements and Predictions, based on Field Failure
Data", Vitro Laboratories, Silver Springs, Md., Report No. 80, October 10, 1955
(Contract NOBSR 63389)
1 8 . D.R . Wal in , "Vacuum Tube Envelope Temperature Measurements" NEL Report #594 ,
March 30 , 1955 , U.S. Navy E l ectron i c s Laboratory , San Diego , Cal i forn i a

1 9 . NEL Rel i ab i l i ty Design Handbook , Section 201. 1 , "Electron Tube Engineering

Data Sheets" , W.B . We l l s , and Section 505.1

20 . "Design Manual of Natural Methods of Cool ing E l ectronic Equipment" , James P .

W e l sh , Cornell Aeronaut ical Laboratory , Buffalo, N.Y. , Report No. HF-845-D-8 ,
November 1 , 1956 (Contract NOBSR-63043)

2 1 . "High Temperature and Al t i tude Li fe Eval u a t i on and Pu l s e Ratings of Submi ni ature

Vacuum Tubes" , Final Engineering Report , Contract AF33(038) 9853, Sylvania
E l ectric Products Inc . , Emporium , Pa.

22. Qual i f icat ion Te s t s of Heat D i s sipat ing El ectron Tube Shields, Test Report # 1 31 ,
January 1956 , United El ectroDynami cs , Divis ion of United Geophysical Corp . ,
Pasadena , Cal i forn i a

23. "Evaluation Tes t s o f Heat D i s s ipating El ectron Tube Shi e l ds o f Various Des i gns"
Test Report #E-472 , March 1957, United E l ectroDynam ics, D i v i s ion o f United
Geophysical Corpor a t ion , Pasadena , Cal i forn i a

24. M i l i tary Spec i f i c a t i ons , El ectron Tubes , MIL-E-l

25. "Evaluat ion Test ing of TR Tube Shi e l d and JAN Bright Tube Shi e l d" Test Report
#E-154 , May 1956, United E l ectroDynam ics Divis ion of United Geophysical Corp. ,
Pas adena , Cal i forn i a

26 . Spec i fication "Heat D i s s ipat ing El ectron Tube Shi e l ds" MIL-S-9372 ( USAF)

27. "Development and Appl icat ion of a Thermal Diode Thermatron for use in Studying
Methods of Cool ing El ectron Tubes" , James P . Welsh , Report No. HF-I053-D-3 ,
May 1957 , Corne l l Aeronau t i cal Laboratory , Buffalo, New York (Contract NOBSR
7253 1 )

28. "Evaluation o f Sh i e l d s for Cool ing M i n i a ture El ectron Tube Type s " , Max B i a l er,
W.A.D . C . Techn ical Report 53- 1 74 , Wright Air Development Center, Wrigh t ­
Pat terson Air Force Base, Ohio

29. Interim t e s t report on Evaluation Tes t s of E l ectron Tube Shi e l ds , May 1 0 , 1956 ,
United El ectroDynam ic D i v i s ion of Uni ted Geophys i cal Corp., Pas adena , C a l i fornia

30. E IA Conference on E l ectronic Equipment Maintainab i l i ty , Univers i ty of Southern

Cal i forn i a , December 18, 1 9 , 1957 . Paper de l i vered by Mr. C.I. Cummings o f
Jet Propul s ion Laboratory , Cal i forn i a Ins t i tute o f Technology.

3 1 . R . D. M i l l s W.W . Wright "The Rating of Thermionic Valves for use under Abnormal
Ambient Condi t ions" Bri t i sh Journal of E l ectronics November 1955

32 . "Qual i f ication Tests of Heat D i s s ipat ing Tube Shields" Report #E-169 , November
1956, Uni ted E l ectrodynamics Divis ion of Uni ted Geophysical Corpora t i on,
Pasadena , Cal i forn i a

United States Patent Office
1 2
2,766,020
ELECTRONIC TUBE CLAMP AND SHIELD
Leroy Ralph Woods, Whittier, Calif., assignor to Inter­
national Electronic Research Corporation, Burbank,
Calif., a corporation of California
Application Augost 28,1953,Serial No. 377,080
9 Claims. (Ct. 257-263)
The invention relates to devices for securing electronic
and similar types of vacuum tubes, which devices serve
also as a means of cooling the tubes by conducting heat
therefrom and also as cushions so as to minimize the
effect of high frequency vibration on the tubes.
Two outstanding essentials in the employment of elec­
tronic tubes have been the provision of tube clamps ca­
pable of securely mounting the tubes in a selected loca­
tion and devices for conducting away from the tubes heat
generated during their operation. New developments in
certain fields and particularly the fields devoted to aircraft
and guided missiles have produced vibration conditions
of far greater amplitude and intensity than previously
experienced. Greater demands upon the performance of
such electronic tubes have also necessitated more efficient
means for conducting heat away from the glass exterior
so as to more uniformly and more thoroughly cool the
tubes.
Spring clips have often been resorted to for holding
tubes in place. These have a number of objectionable
features, one being that spring metal, when heated to
temperatures frequently experienced in this particular
field, tends to lose its resiliency and effectiveness.
It has also been known that glass is a relatively poor
conductor of heat and for this reason tube clamps de­
signed so as to grip the glass at scattered locations fail
as a heat conducting medium in that, although heat may
be conducted very rapidly away from those points on the
glass touched by spring elements or fingers, there is suffi­
cient heat remaining in the glass at locations between
the points gripped by the fingers that excessive strains are
set UP resulting ultimately in cracking and failure of the
tubes. With further respect to the neW and highly inten­
sive vibration conditions it has been found that on many
occasions vibrations set up a resonant condition which
greatly aggravates the objectionable effects normally ex­
perienced in the presence of very high frequency vibra­
tions.
It has further been found that, particularly in flat press
tubes where the area around wires emerging from the
tubes is pressed into l! flat form, normally unequal ex­
pansion between glass and the metal of the wire has
caused a condition making it very necessary to cool tubes
effectivelY at that point.
Efforts to improve the contact areas between the tube
clamp and the glass have met with indifferent success
partly by reason of the fact that relatively large tolerances
are permitted in bulb diameter. Tolerances in surface
areas have also been permitted to the extent that very
frequently seams are present, high spots, and conditions
of out-of-round form sufficient to considerably impair a
close fit of the shield around the tube. This often results
in a gradual diminution of the grip of the clamp on the
tube until, after a period of use but long prior to exhaus­
tion of the tube itself, the grip weakens and conditions
begin to .acc!Jmulate requiring constant servicing.
Good heat conductors can be found which include alu­
minum alloys and spring silver. Alth�ugh both of these
2,766,020
Patented Oct. 9, 1956
metals are highly effective as a heat-conducting medium,
they nevertheless lose their effectiveness when heated due
to losing their temper and consequently their snug grip
upon the tube.
A great deal of trouble has been experienced from
snapping tubes into the spring typ.e clamps. The snap­
ping is sufficient often to start a slight strain not imme­
diately detachable but which after use under vibration
conditions ultimately contributes to fracturing of the tube
10 at an inopportune time.
It is therefore among the objects of the invention to
provide a new and improved tube clamp and heat-con­
ducting shield device which is capable of uniformly cool­
ing an electronic tube on which it may be used.
16 Another object of the invention is to provide a new
and improved tube clamp and heat-conducting shield ca­
pable of operating at a low temperature gradient between
tube and shield.
Still another object oJ the invention is to provide a
20 new and improved construction for a tube clamp and heat­
conducting shield such that the metal best able to conduct
Iieat may be used for that purpose without impairing the
ability of the clamp to maintain a tight grip upon the tube.
Still another object of the invention is to provide a tube
26 clamping means consisting of a relatively dead metal so
associated with a spring clamp that the dead metal is
retained in place securely and uniformly by spring pres­
sure which, because of the absence of heat in the spring
clamp, will not weaken and deteriorate, thereby to di-
30 minish the efficiency of the tube installation.
Still further among the objects of the invention is to
provide a new and improved tube clamp and heat-con­
ducting shield which is especially well adapted to the
clamping and cooling of flat pressed miniature and sub-
35 miniature tubes.
Also included among the objects is to provide a new
and improved cushioned clamp which permits insertion of
the tube without undue strain upon the glass and which
permits removal of the tube from the clamp with equal
(0 ease.
With these and other objects in view, the invention
consists in the construction, arrangement and combina­
tion of the various parts of the device whereby the ob­
jects contemplated are attained, as hereinafter set forth,
46 pointed out in the appended claims and illustrated in the
accompanying drawings.
In the drawings:
Figure 1 is a side elevational view of one form of the
tube clamp and heat-conducting shield device in place
60 upon a flat press tube.
Figure 2 is a vertical end elevational view of Figure I.
Figure 3 is an exploded view showing a tube and
showing parts of the tube clamp and shield device in
the positions from which they are advanced into contact
55 with the tube.
Figure 4 is a cross-sectional view taken on the line
4-4 of Figure I.
Figure 5 is a longitudinal sectional view of a subminia­
ture tube wherein the tube clamp and shield device is
60 ilLustrated with a modified type of
liner.
Figure 6 is a vertical sectional view taken on the line
6-6 of Figure 5.
Figure 7 is a longitudinal view partially in section
showing a form of the deVice applied to a button type
65 tube.
Figure 8 is a fragmentary view of the innermost liner
element employed in the form illustrated in Figures 5
and 6.
In a form of the device chosen to illustrate the in­
'TO vention there is shown a flat press tube 10 conventional
in subminiature type tubes which consists of a cylindrical
portioA 11 and a flat press 12. The flat press produces
 2,'166,020
:3
opposite substantially parallel flat faces 13 and 14 at
the end of which is a transverse bead 15 from which
protmde wires 16.
For holding a tube of the flat press type there are
provided, as illustrated in Figure 1, two principal ele-
ments comprising a chassis 17 and a clamp 18. These
clements are shown in assembled condition in Figures 1
and 2 and separately in Figure 3.
More particularly the chassis is built from a dead metal.
This dead metal is preferably very soft silver initially in
sheet form which may be pressed by means of a die into
the form illustrated in Figures 1, 2 and 3. The chassis
as there shown is provided with protruding configurations
19 spaced longitudinally one from another and extending
substantially throughout the length of the chassis. These
protruding configurations extend across the chassis as
is observable in Figure 4 so as to provide a series of flat
base elements by means of which the chassis is supported
upon some appropriate supporting surface. If the sup­
porting surface be a metal, the chassis may readily be
silver soldered or silver brazed to the base.
The chassis includes a pair of semi-cylindrical side
walls 20 and 21 which can be folded around the cylindrical
portion 11 of the tube after the tube is placed in the
chassis. The semi-cylindrical walls are made of such
width that they almost meet along a longitudinal meet-
ing line 22. It is of interest to note that the protruding
configurations 19 terminate on opposite sides in rounded
ends 23 and 24 which stand out a slight distance from
the exterior wall of the cylindrical portion of the tube.
The configurations may be described as transverse ribs
comprising a bottom support for the chassis.
A t one end of the chassis is a tongue 25 dished to a
limited extent for reception of the end of the cylindrical
portion of the tube where it joins the flat press. The
tongue turns upwardly to a transverse rounded contact
26 which is adapted to have the lower face 14 of the flat
press rest thereon. Ordinarily the end of the chassis
opposite the end at which the tongue is located is open.
After a tube has been set in the chassis and the semi-
cylindrical sides 2(1 and 21 folded around it, the clamp
18 may be applied. The clamp is constructed of spring
material. This clamp has a portion 30 slightly cylindrical
so as to fit the exterior contour of the semi-cylindrical
walls 20 and 21. Along each side of the portion 30 45
is a series of fingers 31 spaced longitudinally from eacn
other and forming therebetween a series of spaces 32.
The fingers continue to carry out the semi-cylindrical
form of the clamp.
At one end of the clamp, namely, the end correspond- 50
ing to the end at which the tongue 25 is located on the
chassis, there are provided a series of spring legs 33
spaced laterally one from another and extending endwise
and downwardly to a location where rounded ends 34 of
the legs are adapted to engage in the face 13 of the flat 55
press.
It will be noted further that the spring fingers 31 also
terminate in rounded ends 35 and that these ends are
adapted to fit between the rounded ends 23 and 24 of the
protruding configurations on the chassis. By fitting thus til)
snugly the fingers prevent endwise shifting of the clamp
relative to the chassis as well as being depended upon to
press the soft metal semi-cylindrical walls 20 and 21
into snug uniform contact with the glass exterior wall of
the tube. G5
Because of the soft character of the semi-cylindrical
walls 20 and 21 the metal of the walls is adapted to and
does conform to irregularities in the exterior surface of
the glass wall of the tube. The fingers therefore serve tolerance allowance in circumference, nevertheless the
as a means of conforming the soft metal of the chassis 70 tube clamp and shield device may be counted upon to
into efficient heat-conducting relationship to the glass function efficiently and successfully. A chassis 46 sub-
of the tube as well as serving to anchor the tube in its stantially similar to the chassis 17 is employed except
bed or position in the chassis. From this description that there is omitted the tongue 25 and its transverse
it will be apparent that the tubes are merely laid in the rounded contact 26. The chassis as previously noted is
soft silver chassis without there being any snapping of 75 constmcted of soft metal, soft silver being especially ap-
the tube in placc which would otherwi3c lend 10 ,,:l lip
a strain in the glass.
After the chassis walls have been wrapped around thc
glass, application of the spring fingers 31, even though
5 they might be snapped into place, does not set up any
strain in the glass of the tube because of the fact that
the soft metallic semi-cylindrical walls 2G and 21 absorb
the shock of snapping the spring fingers into place.
Further by reason of pressing the tongue 25 into snug
10 engagement with the flat press where it is held in snug
engagement hy the legs 33, the flat press is provided with
a means of having heat conducted therefrom which as a
consequence materially improves the cooling of the tube
particularly at the critical point where the wires are em-
15 bedded and emerge.
In a modified form of the invention illustrated in Fig­
ures 5 and 6 there is provided a liner or what may be
aptly described as a cushion liner 36 which may advan­
tageously comprise an initially relatively flat sheet of
20 waffle embossed silver. This is a material frequently in
the neighborhood of about .003 inch in thickness but
wherein the effective thickness due to the waffle grid
effect may be somewhat greater. Details of the cushion
liner are shown in Figure 8.
25 The cushion liner is constructed of such width that it
will substantially surround the cylindrical portion 11 of
the tube leaving only a line 37 of separation.
The positioning of the liner is illustrated to good ad­
vantage in Figures 5 and 6. The length of the liner may
30 be the same or slightly greater than the length of the
cylindrical portion ! 1 of the tube. When applied the
liner is initially wrapped around the tube and then the
tube with the surrounding liner is rested in the chassis as
an object might be rested in a cradle. The semi-cylin­
35 drical sides 20 and 21 of the chassis are then bent around
the cushion liner and ultimately the clamp 18 is applied
over the chassis semi-cylindrical side walls in the same
manner as was initially described in connection with Fig­
ures 1, 2 and 3. This brings the spring fingers 31 into
40 contact with the exterior of the chassis and thereby add­
ing to the force of pressing the liner into conformation
with any irregularities that may exist on the exterior of
the glass wall.
An optional and convenient means of securing the
chassis 17 to an appropriate mounting is illustrated in the
modified form of Figures 6 and 7. For this purpose there
is provided a channel-like clamp or elongated staple 50
having a flat web 51 silver soldered or otherwise at­
tached to the configurations or transverse ribs 19 of the
chassis. At each side of the web is a leg 52, adapted to
be bent as shown by the dotted position 52' in Figure 6.
The clamp thus described is an effective medium for use
in mounting the chassis on one of the commercial forms
of fibre mats in which slots may be cut for reception of
the legs during the mounting operation. The clamp is
one well adapted to mechanical assembly operations.
On those occasions where the tube clamp and shield
device may be called upon to receive and mount a sub­
miniature type tube of the so-called button type, such as
the tube 40 of Figure 7, some slight modification may be
found advantageous. The button type tube usually fea­
tures an annular bead 41 at the pressed end through
which wires 42 dispersed circumferentially about the end
protmde. Frequently there is a cylindrical portion 43 at
the other end of which may be a �ap-like portion �4 ad-
jacent a tip 45.
If the button type tube should have variations in the
exterior contour such as those variations herein sug­
gested or other variations due to roughness or a wide
 2,766,020
5
plicable, so that the metal of the chassis will conform to
the irregularities of the surface of the tube.
A clamp 47 is employed having the same general shape
and configuration as the clamp 18 except that the legs
33 are omitted. Fingers 48 extend along both sides and
are spaced apart as previously noted by spaces 49 there­
between. The fingers fit into the usual spaces between
protruding configurations 19' of the chassis as shown
in Figure 7 and of the same general form as shown and
described in connection with Figures 1 through 6,
inclusive.
Constructed as described the soft metal of the chassis
can sustain a very considerable amount of heat and by
reason of the fact that it can be constructed of the most
efficient heat-conducting metal, the chassis is extremely
efficient in conducting heat away from all sides of the
tube and also from all portions of the tube surfac�.
Continued heating or continued alternate heating and
cooling conditions do not affect the efficiency in any
adverse way. Such effect as might be experienced is
more in the nature of a still closer conformation of the
t:hassis to the glass as the installation continues in use.
When the cushion liner 36 is employed the liner may be
constructed of somewhat thinner material than the chas­
ds can be constructed of and consequently the liner to a
degree even greater than the chassis is adapted to con­
form to such irregularities as may exist due to variations
in tolerance, seams, protrusions and other irregularities
on the tube surface.
Whether the chassis be used alone or in company with
the liner, the soft character of the metal has a very
marked effect in dampening vibrations and destroying
the tendency of the tube clamp to experience resonant
conditions. This is as true of extremely high vibrations
of the supersonic type as in ordinarily high vibrations
heretofore experienced in aircraft and guided missiles.
Uniform pick up of heat from all portions of the glass
virtually eliminates mechanical and heat strains in the
glass, even though portions of the glass may be heated
to considerably higher temperatures than closely adja­
cent porlions. Further by reason of the fact that the
spring fingers of the clamp are not depended upon to con­
tact the hot surface of the glass and further by reason
of the fact that the chassis may be constructed of a ma­
terial having greater heat-conducting qualities than the
siJring clamp, these spring fingers do not absorb a suf­
ficient amount of heat to modify their temper and springy
character. In consequence they maintain their initial
firm grip, holding the sides of the chassis in place despite
extreme conditions in temperature and vibration expected
to be encountered in use. The possibility of the tube
50
loosening from its mounting is therefore materially
minimized.
There has been therefore disclosed herein a highly
efficient tube clamp and shield device which combines
66
the qualities of extremely efficient cooling with vibration­
resistant qualities sufficient that the device may be de­
pended upon to remain securely in place upon any suit­
able face and to securely hold the tube mounted therein
in its proper position.
While I have herein shown and described my inven­
tion in what I have conceived to be the most practical
and preferred embodiment, it is recognized that departures
may be made therefrom within the scope of my invention,
which is not to be limited to the details disclosed herein 65
but is to be accorded the full scope of the claims so as
to embrace any and all equivalent devices.
Having described my invention, what I claim as new
and desire to secure by Letters Patent is:
1. A clamp and heat-conducting shield device for elec­ 70
tronic components comprising a chassis of dead metal
corresponding in length to the length of a component
and having protruding configurations along a bottom
thereof presenting longitudinally spaced transverse ribs
and having substantially semi-cylindrical sides adapted to 76
fold over t he comp onent, and a clamp of spring metai
comprising a semi-cylindrical portion corresponding in
length to the length of the chassis, said clamp having a
series of longitudinally spaced spring fingers on each
IS side of the clamp adapted to extend around the lower
side of the chassis and into spaces between said ribs.
2. A tube clamp and heat-conducting shield device for
electronic fiat press tubes comprising a chassis of dead
metal corresponding in length to the length of a tube and
10 having protruding configurations along a bottom thereOf
and substantially semi-cylindrical sides adapted to fold
over the tube, a tongue at one end of the chassis extending
endwise 1'.nd adapted to engage one side of the flat press
of the tube, and a clamp of spring metal comprising a
lIS dished portion corresponding in length to the length of
the chassis, said clamp having spring elements at the sides
adapted to extend around the lower side of the chassis,
and a spring means extending from one end of the clamp
adapted to engage the other side of said flat press.
20 3. A tube clamp and heat-conducting shield device for
electronic flat press tubes comprising a chassis of dead
metal corresponding in length to the length of a tube
and having protruding configurations along a bottom there­
of presenting longitudinally spaced flat faced transverse
26
ribs and having substantially semi-cylindrical sides adapt­
ed to fold over the tube to a longitudinal meeting line
along the too. a tong ..�e at Oile end of the chassis extending
endwise and adapted to engage one side of the flat press of
the tube, and a clamp of spring metal comprising a semi-
30
cylindrical portion corresponding in length to the length
of the chassis, a series of longitudinally spaced spring
fingers on each side of the clamp adapted to extend
around the lower side of the chassis and into spaces be-
3 tween said ribs, and a spring element extending from
6
one end of the clamp to a position adapted to engage the
other side of said flat press.
4. A tube clamp and heat-conducting shield device
for electroniC flat press tubes comprising a chassis of
40 dead metal corresponding in length to the length of a
tube and having protruding configurations uniformly
spaced along a bottom thereof presenting substantially
flat faced transverse ribs having spaced rounded ends and
having substantially semi-cylindrical sides adapted to fold
over the tube to a longitudinal meeting line along a side
45 remote from the ribs. a tongue at one end of the chassis
on the side adjacent the ribs extending endwise and out­
wardly and terminating in a transverse rounded contact
adapted to engage one side of the flat press of the tube,
and a clamp of spring metal comprising a semi-cylindrical
portion corresponding in length to the length of the
chassis, a series of longitudinally uniformly spaced spring
fingers on each side of the clamp adapted to extend
around the lower side of the chassis and into spaces be­
tween said rounded ends of the configurations, whereby
to press the chassis uniformly into heat-absorbing contact
with the tube, and a series of spring legs extending from
one end of the clamp having rounded ends adapted to
engage the other side of said flat press.
5. A clamp and shield device for electronic components
60
comprising a chassis of dead metal corresponding in
length to the length of a component and having high
heat-conducting characteristics, said chassis having a bot­
tom and semi-cylindrical side walls adapted to extend
around the component, an inner liner of initially flat
dead metal material adapted to lie between the exterior
surface of the component and the interior surface of
said chassis and adapted to extend throughout the cylin­
drical portion of the component, and a clamp of spring
metal comprising a dished portion corresponding in length
to the length of the component, said clamp having spring
elements at the sides adapted to extend partly around the
lower sides of the chassis, whereby said chassis and the
inner liner are pressed into intimate contact with the
exterior wall of the component.
 2,766,020
7 8
6. A clamp and shield device for electronic tubes com­
prising a chassis of dead metal corresponding in length to
the length of a tube and having high heat-conducting
characteristics, said chassis having longitudinally spaced
protruding configurations on one side fmming a bottom
and semi-cylindrical side walls extending to a separation
line along one side, an inner liner of initially flat em­
bossed dead sheet metal material adapted to lie between
the exterior surface of a tube and the interior surface
of said chassis and adapted to extend throughout the
cylindrical portion of the tube, and a clamp of spring
metal comprising a semi-cylindrical portion correspond-
ing in length to the length of the chassis, said clamp hav-
ing a series of longitudinally spaced spring fingers adapted
to extend partly around the lower sides of the chassis
and between the configurations, whereby said chassis and
the inner liner are pressed into intimate contact with the
exterior wall of the tube.
7. A tube clamp and shield device for flat press elec­
tronic tubes comprising a chassis of dead metal corre-
sponding in length to the length of a tube and having
longitudinally spaced transverse configurations on one
side forming a bottom and semi-cylindrical side walls on
t he bottom adapted to extend around the tube, an inner
liner of initially flat dead metal material adapted to lie
betwecn the exterior surface of a tube and the interior
surface of said chassis and adapted to extend throughout
the cylindricalportion .. of.the.tube, and a clamp of spring
metal cOinprising a semi-cylindrical portion correspond-
ing in length to the length of the tube, said clamp having
spring elements adapted to extend partly around the lower
sides of the chassis, whereby said chassis and the inner
liner are pressed into intimate contact with the exterior
wall of the tube, a tongue at one end of the chassis ex­
tending generally endwise into the flat press at one side of
the tube and spring means on a corresponding end o f
the clamp extending generally endwise into the flat press
at the other side of the tube.
8. A clamp and heat-conducting shield device for elec­
tronic components comprising a chassis of dead metal
.It)
having a pocket conforming in general to the size and
shape of the component and having sidc walls of breadth
adapted to be pressed to positions overlying and sub­
stantially surrounding said component, a base-contacting
portion comprising part of said chassis and having an
:i area of the base-contacting portion adapted to be
anchored to the base, and a clamp of spring metal com­
prising a dished portion corresponding in size and di­
mension to the exterior of the chassis, said clamp having
separate longitudinally spaced spring elements at side
10 edges thereof, said spring elements in assembled posi­
�ion extending around the side walls and beneath the
chassis and in resilient contact therewith and separated by
said chassis from direct contact with the component
whereby to urge said chassis into heat dissipating con­ -
Iii tact with the component.
9. A clamp and heat-conducting shield device for elec­
lronic components comprising a chassis of dead metal
corresponding in one lineal dimension to the correspond­
ing dimension of the component and having a plurality
20 of lineally spaced configurations defining engaging areas
for said clamp on each opposite side, said chassis having
sides of breadth adapted to substantially envelope sides
and top portions of said component in an envelope of
said dead metal, and a clamp having one lineal dimension
25 corresponding to the first identified lineal dimension of
said chassis, said clamp having a series of lineally spaced
spring elements on each side integrally connected by a
web portion and resiliently contacting said chassis, said
clamp extending throughout substantially more than half
30 the perimeter of said chassis and adapted to engage said
lineally spaced areas on said chassis whereby to press
said dead metal into engagement with the component.
References Cited in the file of this patent
ai, UNITED STATES PATENTS
2,019,939 Suller _________________ Nov. 5, 1935
2,494,881 Kost __________________ Jan. 17, 1950
2,662,220 Saari __________________ Dec. 8, 1953
2,668,933 Shapiro Feb. 9, 1954
_______________
2,701,866 Chapman ______________ Feb. 8, 1955
Oct. 9, 1956 L. R. WOODS 2,766,020
ELECTRONIC TUBE CLAMP AND SHIELD

Filed Aug. 28, 1953

J�.!I.
� 4--1I
•

DDDDDDD
DODODD
DODDDD
ODODO
�r;pA . IJ
-� •
.

DODD
DO
36

L EROY'RALPH WOODS,
INVENTOR.
HUEBNER, BEEHLER,
WORREL 8 HERZIG.
ATTORNEYS.
♦ Verso Filler Page ♦
 2,807,659
United States Patent Office Patented Sept. 24, 1957

1 2
Another object of the invention is to provide a new
and improved combination tube clamp and shield which
in addition to providing an effective isolation against vi­
2,807,659
brational shock waves is also an effective heat dispersing
TUBE CLAMP A�D SHIELD 5 medium which greatly assists in the dissipation of heat
from the tube under circumstances where overheating
Leroy Ralpb Woods, Wbittier, Cali£', assignor, by mesne might 'be the cause of poor performance, the combined
assignments, to International Electronic i1.c."l','lrdl Cf.'f­
tube clamp and shield being further so constructed that
poration, Burbaok, Calif., a corporation ()� Califomhl
heat is dissipated not only by being reradiated from the
Application March 30, 1953, Scriall'io. 345,597 10 shield but also by being very efficiently conducted through
the shield and its support to the chassis of the installation.
19 Claims. (Cl. 174-35)
Still another object of the invention is to provide a
new and improved tube clamp and shield which very
effectively minimizes microphonics within the tube itself.
The invention relates to the field of electronic ac.::es­ 15 Another object of the invention is to produce a tube
sories · and has particular reference to a de vice for use clamp and shield which has a very appreciable shield­
with conventional vacuum and electronic tubes for the ing effect to protect the tube from stray radiations with­
purpose of holding the tubes in place under circumstances out the necessity of employing a completely enveloping
where there mi gh t be considerable vibration, insufficient metallic envelope.
20
air circulation causing overheating, or perhaps where Also included among the object� of the invention is to
there may be present other electric vibrations or im­ provide a new and improved tube clamp and shield
pulses from which the tube should be shielded. which is extremely inexpensive in view of its high de­
Tube clamps have been in common use for securing gree of lltility. which is substantially universal in its
and protecting vacuum tubes for a great many years, in o� design in that it can be applied to a great variety of dif­
fact, almost as long as vacuum tubes have been in use. _;J ferent types of tube�, and which further is so constructed
The tube clamps heretofore used, however, have been that it can be readily installed upon standard chassis and
little more than relatively thin sheet metal cylinders electro nic installations.
fitting over the tubes in a somewhat careless or indif­ With these -and otht�r obje�i� in view, the invention
ferent .fashion. Various means have been employed consists in the construction, arnln_!!emcnt and c()mhina-
30
for holding the tubes in place but such holding devices tion of the variolls parts of th·� device wJlcreby the oh·
as have been employed have been those which could not jects con te m p late d arc attained, as hereinafter set forth,
be relied upon under all circumstances for properly re­ pointed out in the appended claims and illustrated in the
taining the tube. Under other circumstances where the accompanying dra-,vings.
heating of the tube or cooling of the tube has been a 3:; In the drawings:
problem, tube shields in the fOlm of tube clamps where Figure 1 is an elevation31 sectional view showing the
they have been employed have been little more than tube clamp and shield in place over an electronic tuOC.
stacks for the casual circulatiun of air. Where the Figure 2 is a cross-section taken on the line 2-2
tubes have needed cooling, the tube shields have never of Figure 1.
been so constructed as to provide an effective reradi­ 40 Figure 3 is a fragmentary cross-sectional view taken
ating surface. on the line 3-3 of Figure 1.
More particularly since the employment of immense Figure 4 is an exploded view in perspective showing
numbers of tubes on guided missiles and jet planes, tubes parts of the tube clamp and shield in the form they
have been subjected to vibrations far exceeding uny­ would have prior to assembly.
thing heretofore experienced and also vibrations of a 45 Figllrre 5 is a p erspecti ve view of one of the spring
c haracter not heretofore of great importance. Unu.:r members utilizable in the cylindrical portion of the
thos e circumstances the tube shields which have been cla./np.
established as conventional tube shielding and clamping Figure 6 is an elevational sectional view showing an­
devices have been found to be ineffec t ive in that th.::y other form ('If the device in place over an electronic tube.
permit the elements in the tube to deteriorate very 50 Figure 7 is a cross-sectional view taken on the line
rapidly under the vibrational strains and in som e in­ 7-7 of Figure 6.
stances even to contribute to the breakdown of the ele­ Figure 8 is an exploded perspe ctive view of the holder
ments in the tubes under vibration. On some occa�i()ns and envelope of Figure 6.
where the tubes have been insufficiently and un';atis­ Figure 9 is a perspective view of the spring member
fact orily supported, the vibrational strain betweell the 55 of Figure 6 before being rolled and inserted wit,hin the
tube clamp contacts and the ·glass of the tu be has h�2n envelope.
such as to shatter the glass it�el.f. Tn an embodiment choson to ilh'str:lle Ih.:! iilvention
It is therefore among the objects of the inveroti,n tl') there is shown a t u b e clamp .and shiclct indic:!tcd g·:n­
provide a new and improved tube clamp and �,:I;"U erally by the reference character JO mountcd "pall :1
which is very highly shock resistant to the point v. il':rc GO holder 11 which in turn is mounted upon a support 1 �
the life of conventional vacuum tubes has been kn:;i!l­ which may be a purt of the chassis of some appropri;.t;:
ened by many hundreds of hC'l�rs under extremely ;.d­ electronic unit. A socket 13 of any <'nt! of the seve r a l
verse circumstances. conventional types is mOllnteu beneath the support J2
Another object of the invei;:ion is to prov i d e a nl'w in a scmewhat loose f:lshitln as is customary in the trad.:.
and improved tube clamp and ',hdJ whi-.:h is ob , c ; lii,' I;;:;
' The holder 11, shown in perspective view in FigUrre
tube with an extremely high i.h.:grce of effectivencs�; by 4, consists of a disc-like annular r i ng 14 p rov ided with
providing a great many separate and individinl resilient lugs 15 having holes 16 therein by means of which the
contacts between the tube clamp and tile glass of the holder can be secured to the support 12 by means of
tube on all sides and throughout its length so that vi­ conventional screws. The annular ring portion has a
brational shocks are evenly distributed and dampened 70 series of outwardly extending retainers 17, four in num­
to the extent that they become almost ineffective to cause ber, having overhanging lips 18 providing slots 19. At
deterioration of the tube. the inner edge of the annular ring is an upstanding flange
 2,807,669
3 4
20 from which a series of four friction guides 21 extend the holder 11 upon the support 12 by me a ns of suitable
up wardly , as viewed in Figure 4. screws 39, as illustrated in F igur e 2. Wi th the holder in
The main po rt io n of the tube clamp and shield consists place the tube 27 is then i n sta l led and after its installa­
of a metal envel o pe 22 which may, for example, be of tion the tube clamp is placed over the tube. The feet 28
thin �heet aluminum, the envelope as herein illustrated 5 initi ally fall into spaces between the retainers 17 with a
being uniformly cylindrical and having an open top 23 lowermost ring portion 40 at the base of the envelope
and an open bottom 24. The top for convenience may fitting with a snug friction fit around the exterior of the
be provided with an o verhang ing flange 25 a dapt ed to flange 20 and within the innermost faces of the lips 18.
retain bene�lth it a s pring 26 of progressively diminishing The c lam p is then rotated until the feet 28 slide with a
diameter, as illustrated in Figure I, so th at the outcrmo�t 10 snug friclion fit into the respective slots 19. In this posi­
turn of the spring can be secured beneath th.: Ibng;.: and tion the S;)I ing 26, if a spring be used, will press against
the innermost turn, of smal lest diameter, be adapted t o the top of the tube 27 and positively prevent its dislodge­
press r e s ilient l y against the top o f a vacuum tube 27. ment. At t he same time virtually eaeh and everyone
At th e base of the envelope is a s eries of four outwardly of the s pr ing elements 34 and 35 press each in its turn
ex tending feet 28, these feet being of a propcr thickne;,s ,:j ;:gainst the re<;pective section of the exterior surface of
and l en g th so as to sl ide snugly within the sl ot s 19. the tube 27 either on the glass portion or at its base. Th e
It will be not ed that the envel o p e 22 is pr o v i ded with tension of the spring elements can be adjusted so that it
a series of four substantially rectangular openings 29 will b e proper for anyone of a great many different cir­
spaced e v e n l y around the circumference of the envelope. cumstances where different vibrational shock waves may
Between the o pe ning s 29 arc imperforate portions 30 of 20 be encountered. The tension may be relatively light at
substantially equal width and length. The openings 29 each resilient contact but by reason of the fact that there
extend t hrou g h o u t a di s tan c e almost the height of the are a multiplicity of contacts distributed in a uniform
en ve lo p� and are slightly greater in width than the width well-defined pattern ove r the lengt h and circumference of
of the intervening imperforate por tions 30. the tube, the total frictional support for the tube will be
Spring members 31, illustrated in Fi g ur e 5, are four in �5 ap p reciable and sutlicient to hold the tube in place even
number and provide the direct support for the tube 27 in the absence of the spring 26. M oreo ve r , by uniformly
within the t\1be clamp. Each spring member is identical supporting the tube in this manner by a multiplicity of
and consists of a ce n t ral column or rib 32 , the relative resilient spring elements distributed as described, vibra­
length of which can be readilv determined from an in­ tional shocks will be almost completely absorbed in the
spection of Fi gu re 1. The c �lumn is ordinarily made :�o spring elements and not conveyed to the tube. It may be
from a piece of ,heet stock preferably spring co ppe r which said that the tube fl o ats within the clamp i s ola t ed against
l·:tn be formed by a suitable set of d i es into the shape vibration.
illustrated in Figure 5. As there shown on each longitu­ Further, by reason of the fact that th e spring el em en ts
dinal edge 33 is a ser ies of laterally e xtending leaf spri n g 34 extend into the openings 29, there is a grid-like pattern
elements 3.J which extend in a gen erally circumfe rent ial ly .,,j effect entirely s u rrounding the tube with such sp acing
arcuate d i re c t i on . as cl e ar ly shown in Figure s 2 and 3. between the mctallic e l ement s that the clamp forms an
Alll!rn;!!i'1� wit h the leaf spring elements 34 is a series elfectiv e shield against stray radiation. Still further, by
of rev.:rscly bent i n w ardl y directed lea f s p r ing elements reason of the fact that the combination of spring elements
35 which alternate with tbe leaf spring elements 34 on 35 lie in open spaces, such as the openings 29, free cir-
both of thc 10ngitLldinal edges 33 of the central column ·10 eulation of air is assured in and around the tube and the
or rib 32. The sp ring elements 35 also t a k c a cin; u m­ amhient temperature can be controlled very effectively
fc:rcl1!i:lIly arcuate direction, as clearly ill ll str ated in Fig. using suitab le conventional air circulators. At the same
2. �o that both the elements 34 and 35 conform to the time the spring elements provide an open const ruction
a rcuat e exterior of the glass of the tube 27. Thc spring permissive of free circulation of air and moreover supply
membcrs 31 may be secured each to its respective i m per ­ 45 a multitude of heat-conducting contacts from the glass
(orate portion 30 by means of rivets 36, f or example, as to the holder and the s u pp or t . A substantial q uantity of
illustrated in Figure I. Rivet holes 37 arc pro\'ided in the heat from th e tube will thereby be dissipated by con­
the central column 32 . duction. Heat conducted to the spring elements and the
Onc spring member being thus se cure d at each one of envelope can to a c onside rab l e degree be also dissipated
the imperforate po r t ions 30 provi des a spring m e mber on ;)0 by reradia t ion from th e roughened and darkened finish
each of four sides of the tube 27 h av i ng a position en­ sur face s . J t has been fou n d advantageous to employ an
:.tblin g the sp ring members to dist r ibute resilient pressure envelope black on both sides, the �urface of which may
e ve nly on all siJes of the tube. be r o ugh ened either by sand-blasting or a wrinkle finish.
For additional safety and secur i ty there is provided at The outside of the spring e l eme n ts may be similarly
the OO l l om of each spring clement 31 a resilient poc k et , ;).j finished.
\\hil.'h pocket is constructed of an inwardly dished por­ In an other form of the device illustrated in F igures 6
tion 38. i l l u st rated in Figure; 3 and 5. The res ilient through 9, in clusive , the device is shown in a more sim­
pocket is adapted to f rictionally receive th e re specti ve plified construction wh erein the envelope and spring mem ­
friction guide 2t already d esc ribed as being a part of the ber are built from sheet metal rolled to cylindrical form
Ilange 20 of Ihe holder 11. {l0 and contained one within another and within a suitable
In the construction of the tube clamp and shield it has holder at the bottom.
been found advantageous to employ a black cadmium More pa rt icul a rl y , there is provided a h older 50 annular
finish on the spring members. The envelope 22 c an more in sh ape and provided with a base flange 51 adapted to be
advantageously be constructed with a black crackle finish, mounted upon the supporting surface 12 immediately
both finishes being such as to greatly enhance the dissipa­ (i5 above the sock et 13 as has b een previously described in
tion of heat by rad iation . Inner faces of the sprint; ele­ wnnection with the holder It. Screws 52 are sug­
ments 34 and 35 should carefully conform to the shape gested for attaching t he holder to the supporting surface.
of the glass of the tube and be cleaned sufficiently to pro­ The holder includes an u pstand ing ring portion 53 at
vide a good contact. It is also very greatly advantageous the top of which arc ears 54 which have inwardly turned
to provide an effective elec t ri c and heat conducting con­ 70 ponions 55 all sloping in the same general direction.
tact between the feet 28 and the reta iners 17. The good An envelope 56 is con stru c te d of s h ee t metal rolled into
contact is further enhanced by a close metal to metal fit the cylindrical form, illustrated in Figures 6 and 8, such
of the friction guides It, each in its respective rc�ilient that one 10tl,-:i!l:Jil1�tl cdge 57 takes a position in abutment
pocket. with the other longitudinal edge 57'. The envelope is
In use of the tube clamp it is customary first to mount 75 provided with large cut-out areas 58, 58', 58" and 58"'.
 2,807,669
5 6
The upper end of the envelope shows a continuous an­
nular portion 59 except for the split between the edges
57 and 57'. The envelope is ordinarily rolled to a di­
mension so that the outer diameter is slightly greater than
the inner diameter of the ring portion 53 so that it may
be contracted when mounted on the holder and there­
after permitted to spring outwardly into snug engage­
ment with the holder.
For connecting the envelope 56 to the holder 50 there
are provided obliquely disposed slots 60 at the base of
the envelope, these having a width adapted to receive the
il�wardly turned portions 55 of the ears 54. By provid­
ing longitudinal openings 61 the inwardly turned por­
tions 55 are udmitted 10 the location of the slots 60 and
the envelope can thereafter be rotated so as to force the
inwardly turned portions 55 into the lowermost ends of
the slots.
To maintain the envelope in its cylindrical form there
is provided at the upper end an annular continuous flanged
retainer 61 '. This is preferably constructed with an in­
wardly directed flange 62 overlying the top and serving
a:, a keeper for a conical coiled spring 63.
A vertical side wall 64 of the retainer extends down­
wardly overlying the exterior of the envelope but more
particularly extending down to a location slightly lower
than the annular portion 59 so that the side wall 64 may
be crimped at points 65 so as to underlie the annular
portion 59 at the tops of the cut-out areas 58, 58', etc.
A spring member 66 is shown in Figures 6 and 7 lo­
cated within the envelope 56. The spring member is
shown in sheet form in Figure 9. As suggested the spring
member is constructed from sheet metal having suitable
resilient characteristics to support a multiplicity of curved
spring elements 67 on one face thereof, this being the in­
ner face when the spring member is rolled into the cylin­ 35
drical form illustrated in Figures 6 and 7.
The spring member may have the spring elements cut
Ollt of the material of the member and turned into some­
what curved shapes inwardly forming some selected pre­
determined pattern of spring elements over substantially 40
the entire surface of the spring member which will en­
compass the electronic tube. While the particular pattern
is not critical, it is, however, important that there be a
somewhat uniform distribution of spring points so that
the resilient pressure on the tube will be distributed to 45
those portions of the tube which need supnort in order
that there be balanced support on all sides arranged w
as to most effectively dampen vibrations ,· hich might
..
exist.
The spring member 66, being inherently resilient in 50
charactcr, is preferably rolled and then inserted within
the envelope 56. Outwardly pressed lugs 68 at the bottom
readily protrude outwardly into th e adjacent cut-out areas
58 and similar lugs 69 :It the top also protrude outwardly
into the respective cut-out areas. By spacing the IllgS
68 and 69 at proper distances longitudinally, they serve
to engage the underside of the annular portion 59 at the
top and the upper side of a corresponding portion at the
bottom so that once in place the spring member will not
shift endwise. As previously noted, the inherent resili­
ency will expand the spring member outwardly into snug
engagement with the inside wall of the envelope.
The coiled spring 63 may be relied upon to additionally
secure the electronic tube in place and to assist in properly
holding the tube clamp in its surrounding position.
There h,IS thus been described a tube clamp and shield
of extremely simple character but which by reason of
it!> constrtlction is highly effective as a shock minimizer,
heat radi:; tor, and shield. The construction, moreover,
i� one of extremely simple character capable of being
formed hy relatively inexpensive dies and machinery to
the end th :[! tube clamp� of the sort herein described can
be provid·:d in ample quantities to fit practically any
in�tallatiu!l.
While r have herein shown and described my invention 75
in what I have conceived to be the most practical and
preferred embodiment, it is recognized that departures
may be made therefrom within the scope of my inven­
tion, which is not to be limited to the details disclosed
5 herein but is to be accorded the full scope of the claims
so as to embrace any and all equivalent devices.
Having described my invention herein, wha t I claim
as new and desire to secure by Letters Patent is:
1. A tube clamp and shield for electronic tubes com-
10 prising an encircling metal envelope, means at the base
ef the envelope for attaching the envelope to a holder,
and means for holding the tube within the envelope com­
prising a multiplicity of spring contact elements of heat
conducting metal on the inner face of the envelope ar-
15 ranged in a uniform pattern throughout the inner circum­
ference and length of the envelope, said contact elements
being spaced one from another in a circumferential direc­
tion and being spaced one from another in a direction
transverse to the circumferential direction and being
:W adapted to resiliently engage the exterior of the tube.
2. A tube clamp and shield for electronic tubes com­
prising a circumferentially continuous metal envelope,
means at the base of the envelope for attaching the en­
velope to a holder, and means for holding the tube within
25 the envelope comprising a multiplicity of spring contact
elements of heat conducting metal on the inner face of the
envelope spaced one from another in a longitudinal di­
rection forming rows, said rows being arranged at lat­
erally spaced intervals throughout the inner circumference
30 of the envelope and adapted to resiliently engage the
exterior of the tube.
3. A tube clamp and shield for electronic tubes com­
prising a substantially cylindrical metal envelope greater
in length than the length of the tube and greater in diam-
eter than the di:Jmeter of the tube, a configuration at the
base of the envelope comprising means for attaching the
envelope to a holder, and means for holding the tube
within the envelope comprising a multiplicity of sub­
stantially flat surfaced spring contact elements of heat
conducting metal on the inner face of the envelope spaced
one from another in a longitudinal direction forming
rows, said rows being spaced laterally and numbering at
least eight rows of alternately spaced elements adapted
to resiliently engage the exterior of the tube and extend-
ing uniformly throughout the inner circumference of the
envelope.
4. A tube clamp and shield for electronic tubes com­
prising an encircling metal envelope having an open base
and clamping means at the base adapted to seCUre the
envelope in a holder, said envelope having a series of
elongated circumferentially disposed openings separated
by portions of the envelope, and a spring member on
said portion having longitudinal edges and comprising a
series of longitudinally spaced inuividual laterally extend-
55 ing spring elements of heat conducting metal on each
longitudinal edge, and a second series of individual longi­
tudinally spaced spring elements on each longitudinal
edge extending across the said portion of the envelope,
all said spring elements being biased toward the interior
60 of the envelope and adapted to press into firm contact
with said tube simultaneously on all sides thereof.
5. A tube clamp and shield for electronic tubes com­
prising a circumferentially continuous metal envelope
having an open base and clamping means at the base
65 adapted ti) s,-�lIre the envelope in a holder, said envelope
having a series of elongated circumferentially disposed
openings sq�arated by integral portions of the envelope,
and a sep:'rate spring member mounted on each integral
portion comprising a series of longitudinally spaced in-
70 dividual laterally extending spring elements on each
longitudi 11:11 edge adapted to be positioned in the respec­
tive opening at the near side thereof and a second series
of indiviuclal spring elements of heat conducting metal
on each Icngitudinal edge extending across said portion
of the envelope, said spring elements being biased toward
 2,807,669
7 8
the interior of the envelope and adapted to press into firm
contact with said tube simultaneously on all sides thereof.
6. A tube clamp and shield for electronic tubes com­
prising a substantially cylindrical metal envelope having
an open top and base and clamping means at the base
adapted to secure the envelope to a holder, said envelope
having a series of four elongated circumferentially dis­
posed openings separated by integral portions of the
envelope. and a separate spring member mounted on
each integral portion comprising a series of individual
laterally extending leaf spring elements on each longi­
tudinal edge at uniformly spaced distances apart adapted
to be positioned in the respective opening at the near
side of said opening and a series of individual reversely
inwardly bent leaf spring elements of heat conducting
metal on each longitudinal edge alternating with the first
identified spring elements. all said spring elements having
inside faces thereof curved in conformity with the exterior
of the tube and adapted to press into firm contact with
said tube simultaneously on all sides thereof.
7. A tube clamp and shield for electronic tubes com­
prising an annular holder, a substantially cylindrical
metal envelope, means on said holder and at one end of
the envelope forming a releasable connection there­
between. said envelope having longitudinal openings 2,) portions of the envelope between t h e openings a n d hav­
therethrough. and an outwardly expansible spring mem­
ber comprising a substantially cylindrical sleeve of sheet
metal adapted to be contained in said envelope with an
outwardly springing fit and interlocking means respec­
tively on said envelope and said sleeve, said sleeve having ;W the length and circumference thereof, a series of cir­
a multiplicity of inwardly curved resilient spring portions
of heat conducting metal scattered in a predetermined
pattern around the interior surface and adapted to engage
the tube at a corresponding number of locations_
8. A tube clamp and shield for electronic tubes corn-
prising an annular holder, a metal envelope of sheet
metal formed into a cylinder with longitudinal edges
adjacent each other, said envelope having an external
diameter greater than the internal diameter of the holder
and adapted to be compressed and to spring outwardly
into contact therewith, means on said holder and at onl1
end of the envelope forming a releasable connection
therebetween. said envelope having longitudinal opening�
therethrough, an expansible spring member comprising
a substantially cylindrical sleeve adapted to be restrained
by and contained in said envelope and interlocking means
respectively on said envelope and said sleeve. said sleeve
having a multiplicity of inwardly curved resilient portions
of heat conducting metal scattered in a predetermined
patlcrn around the interior surface and adapted to engage .,)0 mally adjacent the envelope, and means at the base of
the tube at a correspondin!! number of locations.
9. A tube clamp and shield for electronic tubes com­
prising an annular holder, a metal envelope of sheet metal
formed into a cylinder with longitudinal edges adjacent
each other, said envelope having an external diameter .i:i
greater than the internal diameter of the holder and
adapted to be compressed and to spring outwardly into
contact therewith, means on said holder and at one end
of the envelope forming a releasable connection there­
between, said envelope having longitudinal openings
therethrough, an expansible spring member comprisin:;
a substantially cylindrical sleeve of sheet metal adapted
to be contained in said envelope with an outwardiy
�pringing fit at a location spaced from the connection of
the cnvdope with the holder and interlod.ing mealls
respectivdy on said envelope and said sleeve, ,aid sleeve
having a Illultiplicity of inwardly curved resilient portiom
of heat eondllcting metal scattered in a predetermined
patlern around the interior surface and adapted to engage
the Illhc a! a corresponding number of locations.
10. A tube clamp and shield for electronic tubes com­
prising an annular holder, a metal envelope of sheet metal
formed into a cylinder with longitudinal edges adjacenl
each other, said envelope having an external diameter
greater than the internal diameter of the holder and
adapted to be compressed and to spring outwardly into
contact therewith, means on said holder and at one end
of the envelope forming a releasable connection there-
between, said envelope having longitudinal openings
5 therethrough, an expansible spring member comprising a
substantially cylindrical sleeve of sheet metal adapted to
be contained in said envelope with an outwardly springing
fit and interlocking means respectively on said envelope
and said sleeve, said sleeve having a multiplicity of
10 inwardly cllrved resilient spring portions of heat conduct­
ing m;?u;1 seatiered in a predetermined pattern around
the interior surface and adapted to engage the tube at a
corresponding number of locations, said sleeve being
located at a position spaced outwardly of the connection
\.') bc(\\cen the holder and the envelope, and an annular
con,; :i,i(l!!, fLnged retainer ring secured at the top of the
envelope adapted to fix the circumference of the envelope
and to provide a coil spring keeper.
I L A tube damp and heat disperser comprising a
20 substantially cylindrical envelope having heat radiating
surfaces and having an open base, a holder adapted for
mounting on a support means on the base of the envelope
adapted to engage the holder, circumferentially spaced
elongated openings in the envelope and spring means o n
ing a multiplicity of substantially uniformly spaced indi­
vidual spring elements of heat conducting metal form­
ing a pattern of resilient contacts at the interior of the
envelope and adapted to engage the tube throughout
cumferentially fpaced friction guides extending from the
edge of the holder normally adjacent the envelope and
resilient means at the base of the envelope forming a
resilient connection between the envelope and the holder.
33 I:!. A tube elamp and heat disperser comprising a
circumferentially continuous envelope having heat radiat­
ing surfaces and having an open top and base, a holder
adapted for mounting on a support having overlying lips
forming slots therebeneath, and feet on the base of the
-lO envelope adapted to engage within the slots, circumfer­
entially spaced elongated openings in the envelope and
:spring means on portions of the envelope between the
openings, said spring means comprising a multiplicity of
substantially uniformly spaced individual spring elements
.i;"i of heat conducting metal forming a uniform pattern of
resilient contacts at the interior of the envelope and
adapted to engage the tube throughout the length and
drcumference thereof, a series of circumferentially spaced
friction guides extending from an edge of the holder nor-
the envelope comprising adjacent portions of the spring
means and the envelope forming a resilient pocket lldapted
10 receive the guides when the envelope is in place on
the holder.
13. A vibration damping and heat dissipating clamp
and shield for an electronic component comprising the
combination of an encircling metal envelope exceeding in
length and perimeter the length and perimeter of said
component and ad::1pted for attachment to a base, and
00 an insert for said envelope comprising a member of resili­
ent heat conducting metal extending throughout the in­
terior of said envelope and adapted to be confined therein,
said insert including a multiplicity of inwardly extending
separate spring elements of substantially uniform size,
ti5 said spring elements being spaced substantially uniformly
in both longitudinal and circumferential directiol1� and
"dapted to engage the component at a corresponding num­
ber of locai ions whereby to m a int ain said insert at a loca­
tion spaced uniformly from said component, said insert
7t) having an outwardly springing engagement with the in­
terior of said envelope.
14. A vibration damping and heat dissipating insert for
d'�mping and .·.hielding an electronic component and
"dJptcd for retention in a metal envelope, sait! insert com­
i;') prising a member of resilient heat conductnig metal hav-
 2,807,659
9 10
ing a length and perimeter exceeding the length and engage corresponding arcuate areas on the exterior o f the
perimeter of the component, said insert including a multi­ component and to establish a pattern of resilient engage­
plicity of separate spring elements extending inwardly ment areas throughout the !:::ngth and perimeter of the
therefrom at uniform distances, said spring elements being tube, and outwardly expandJole ::Ireas on the exterior of
spaced substantially uniformly in both longitudinal and 5 said liner located intermediate said arcuate areas and
circumferential directions and adapted to engage the com­ forming a pattern of envelope engagement areas.
ponent at a corresponding number of locations whereby t o 1 8 . A vibration damping and heat dissipating clamp
maintain said insert at a location spaced uniformly from and shield for an electronic component comprising the
said component. combination of an encircling metal envelope adapted
15. A vibration damping and heat dissip:!ting insert for 10 for attachment to a base, an insert for said envelop.:
clamping and shielding an electronic component and comprbing a Illember of resilient heat conducting she;'!t
adapted for retention in a metal envelope comprising a metal, s<Jid insert including :l multiplicity of spaced spring
member of resilient heat conducting sheet metal having elements on said insert extending inwardly and through­
a length and perimeter exceeding the length and perimeter out the length and perimeter of the insert, said spring
of the component, said insert having a multiplicity of 15 ekments having innermost portions of substantial length
separate spring elements comprising cut-out portions of �nd breadlh forming substantially arcuate areas adapted
said member and bent inwardly therefrom and termi­ to engage corres[1onding arcuate areas on the exterior of
nating at uniform distances from said member, said spril!� the component and to establish a pattern of resilient en­
elements being productive of contact areas of substantial gagement areas throughout the length and perimeter of
uniforlllity throughout both the length and perimeter of 20 the tube, and areas on the exterior of said liner inter­
the component. mediate said arcuate areas in resilient heat conducting
16. An electronic tube clamping and shielding insert engagement with the interior of said envelope.
for a tube shield envelope comprising a sheet of m�t.:rial 19. A vibration damping and heating dissipating insert
in the form of a resilient cylinder substantially equiva­ for clamping and shielding an electronic component and
lent in circumference and length to the exterior of the 25 adapted for retention and mounting by means of a metal
tube, said cylinder h"-ving a multiplicity of spring contact envelope comprising a member of resilient heat conduct­
e1eml?llts of heat conducting metal extending inwardly ing metal, said insert including a multiplicity of spaced
of the inner face of the cylinder, said elements being spring element contacts on said insert extending inwardly
spaced circumferentially and longitudinally at substan­ and throughout the length and per i m e t e r of the insert,
tially equal intervals and each element including a tube 30 said spring elements having innermost portion<; of sub­
contacting area having a length and breadth conforming stantiallcngth and breadth forming a substantially arcuate
to an ar:::a on the tube of corresponding length and area pattern adapted 10 engage a corresponding arcuate
breadth. said elements being adapted to surround the area pattern on the exterior of the component and to estab­
tu be throughout the length and circumference thereof, lish a resilient engagement area pattern throughout the
said spring contact elements having faces all at substan- 35 length and perimeter of the component, and means form­
tially the same distance inwardly of the envelope whereby ing an opposite outwardly expandable spring contact area
to provide a multiplicity of combined heat transfer and pattern on the exterior of said liner adapted to form an
spring contact areas between the tube and the envelope envelope engagement area pattern whereby to transfer
throughout the length and circumference of the tube, said heat from the component to the envelope and to dampen
insert when confined in the envelope having portions 40 vibration between the component and the envelope.
thereof adapted to resiliently engage the tube in an in­
ward radial direction and to engage the envelope in a References Cited in the file of this patent
radially outward direction whereby to provide spring pres­ UNITED STATES PATENTS
sure simultaneously against the tube and the surrounding
45 1,426,861 Haddock Aug. 22, 1922
envelope.
______________

2,398,626 Del Camp Apr. 16, 1946

17. A vibration damping and heat dissipating insert
____________

2,472,747 Jones June 7, 1949

for clamping and shielding an electronic component and
_________________

2,512,424 GoJdstaub June 20, 1950

adapted for retention in a metal envelope comprising a
___ __________

2,646,460 DeI Camp July 21, 1953

member of resilient heat conducting sheet metal, said
_____________

2,745,895 Lideen �ay 15, 1956

insert including a multiplicity of spaced spring elements 5 0
_______________

on said insert extending inwardly and throughout the

FOREIGN PATENTS
length and perimeter of the insert, said spring elements
h,aving innermost portions of substantial length, and 350,471 Great Britain ___________ June 4, 1931
breadth forming substantially arcuate areas adapted to 584,310 Germany _____________ Sept. 18, 1933
Sept. 24, 1957 L. R. WOODS 2,807,659
TUBE CLAMP AND SHIELD

Filed larch 30. 1953 2 Sheets-Sheet 1

.It)
�O rpiIiI��'W
,,19
.10 .17

LEROY RALPH WOODS,

INYENTOR.
HUEBNER, BEEHLER,
WORREL a HERZIG,
BY! ) . ATTORNEYS.
� ��.�If
�.....-
Sept. 24, 1957 L R. WOODS 2,807,659
TUBE CLAMP AND SHIELD

Filed March 30, 1953 2 Sheets-Sheet 2

LEROY RALPH WOODS,

INVENTOR.

HUE.BNER, BEEHLER,
WORRE L 8 HERZ IG,
ATTORNEYS.
BY
U� t;) QQJL P-I-z.-
♦ Verso Filler Page ♦

United States Patetlt Office
1 2
2,899,670
TUBE SHIELD ATTACHMENT
Leroy R. Woods, Whittier, Calif., assignor to International
Electronic Research Corporation, Bnrb:mk, Calif., a
corporation of California
Application February 6, 1956, Serial No. 563,655
3 Claims. (Cl. 339-143)
The invention relates to shielding holders for electronic
components and has particular reference to an attachment
by means of which the casing of a tube shield is attached
to a base, the base being of some appropriate type in turn
secured to the chassis on which the mechanism is mounted.
There has been considerable impetus within the past
few years in connection with the improvement of tube
shields of a type found particularly useful for the mount­
ing of electronic tubes. More exacting performance re­
quirements have forced the attention of tube shield manu­
facturers on the design and production of tube shields ca­
pable of more effectively shielding a mass of tubes in the
electronic installation and also capable of easy attachment
and removal without sacrificing effective performance.
Tube shields heretofore employed, while bei!lg equipped
with dependable attachment means, have been of such
construction that the securing of the shield to the base
in many instances is greatly impaired by the crowded
condition within an installation or on occasions have been
such that once installed and left in use they become dif­
ficult to remove without a partial dismantling of the sys­
tem. This has given rise to servicing problems which have
become a burdensome condition.
Tt is therefore among the objects of the invention to
provide a new and improved tube shield attachment which
is simple in construction and inexpensive to manufacture
and install.
Another object of the invention is to provide a new and
imp.oved tube shield attachment which by reason of its
structure and assembly is one, the parts of which cannot
be easily damaged to the ext'!nt that employment of the
attachment becomes difficult.
Still another object of the invention is to provide a new
llnd improved tube shield attachment whereby the tube
shield casing can for all practical purposes be easily ap­
plied to the base at any position of rotation of the casing
with respect to the base without it being necessary for
the service man to hunt for the proper position of one with
respect to the other before the attachment is locked.
A still further object of the invention is to provide a
new and improved tube shield casing and base combina­
tion which make firm engagement with each other to the
end that the casing will not vibrate loose from the base
even under conditicntions of severe vibration and which
moreover makes a sufficiently tight and complete metal to
metal contact that heat generated in the electronic com­
ponent can be readily passed by conduction through the
casing to the socket and thence dissipated in the chassis.
With these and other objects in view, the invention con­
sits in the construction, arrangement and combination of
the various parts of the device whereby the objects con­
templated are attained, as hereinafter set forth, pointed
out in the appended claims and illustrated in the accom­
panying drawings.
In the drawings:
Figure ! is a side elevational view of the shield casing
and base in assembled form shown in perspective.
2,899,670
Patented Aug. 11, 1959
Figure 2 is a longitudinal sectional view ef the com­
bination.
Figure 3 is a cross-sectional view taken on the line 3-3
of Figure 2.
5 In the embodiment of the invention chosen for the pur-
pose of illustration there is shown a tube shield casing 1;)
mounted upon a base 11 which in turn is supported upon
a chassis 12. A glass envelope 13 of an electronic tube
is shown mounted on the base.
10 More particularly the base consists of a shoe 14 hav-
ing ears 15 extending outwardly therefrom in a position
adapt·�d to lie flat upon the chassis where the base can be
�,:,cured by appropriate screws 16. The specific means of
attachment of the base to the chassis will of course depend
15 upon particular conditions encountered in a given instal­
lation. A socket 17 forms part of the base and within the
s{)c!,et is an appropriate plug of a substantially conven­
tional sort adapted to receive the base of any Oi1e of a
great variety of conventional electronic tubes. Oil onc
20 sitle side of the socket and near the outer end thereof is
a projecting configuration 18 which for convenience and
economy can be made of a pressed out portion of the
material of the socket 17. In the chosen embodimC!lit the
configuration has a button-like appearance arcuate at the
25 base and having a height preferably very slightly exceed­
ing the thickness of the casing i O.
The casing 13 in the chosen embodiment has a some­
what elongated cylindri-;al wall i9, an outer end of which
includes an inwardly extending annul"r flange 2' providing
30 a central opening 21. As noted the length of the casing
is somewhat greater than the over-all height of the glass
envelope so that if need be a spring 22 may be employed
having its largest coil retained beneath the flange 23 ailtl
its smallest coil bearing downwardly upon the upper end
35 of the giass envelope. The diameter of the inner circum­
ference of the wall 19 is slightly larger than the exterior c.f
glass envelope, thereby to provide an annular space 23
between the glass envelope and the wall.
An inner end 25 of the casing has an inside diameter
40 the same as or perhaps slightly smaller than the outside
diameter of the socket 17. A snug sliding fit between
the inner end 25 of the casing and the socket is made
possible by providing in the ccsing a plurality of slots :;6.
In the chosen embodiment five such slots have been used.
45 Each slot is identical cnd has oblhpe walis 27 at the
open end of the slot providing an entrance opening 2lS
somewhat wider at the mouth than tile greatest dimension
at the base of the configuration lit Intermediate ends
of each of the slots 26 is an enlargement 29, the circl!m-
50 ference of which is substantially equal to the circumfer­
ence at the base of the configuration 18. The center of
each of the enlargements 29 is located a distance from
the inner end ZS of the casing such that when the con­
figuration is received in the enlargement, the casing will
55 be in its proper preferred position on the base.
By making the slots 26 sufficiently long as suggested in
the drawing, a springiness is provided in the areas of ma­
terial between the slots so that these areas can spread
slightly when the casing is pressed over the soc!;.et. A
60 large area of contact, however, is advantageous and for
that reason there is not provided any marked difference in
diameters between the inner wall of the casing and the
exterior wall of the socket.
When the casing is applied to the base, it need be
(i5 rotated only slightly until some portion of the entrance
c-pening 28 lies adjacent the configuration 18. Further
pressure upon the casing will cause it to rotate slightly
until the configuration is in alignment with the slot 26,
at which point the immediatelY adjacent areas at the side
of the slot 26 may spread slightly to permit passage of
70
the configuration 18 until the configuration is received
within the enlargement 29. At this point the adjacent
 2,899,670

3 4
areas wiII contract by reason of their inherent resiliency casing having a socket-engaging open end adapted to
and the configuration wiII thereafter retain the casing fit over said socket member with a snug sliding fit, said
upon the socket and the base. Only one configuration open end having a plurality of circumferentially spaced
is necessary by reason of an appropriate fit of the casing slots extending inwardly from the end and forming there-
around the socket. When the casing is to be removed it 5 between a series of elongated resilient members, said slots
is necessary only to pull endwise on the casing, in which having enlargements comparable in size to said con­
event the configuration wiII again spread the areas at figuration intermediate outer and inner ends whereby
the sides of the respective slot, thereby permitting the when said casing member is forced over the socket mem­
casing to be withdrawn. ber engagement of one of said slots with said configura-
It will be apparent from the description that hunting 10 tion is adapted to spread adjacent resilient members and
for the proper connection between the casing and the to guide said casing member to a position wherein said
base is minimized by providing only one configuration on configuration projects into the enlargement of the respec­
the base and a multiplicity of slots on the casing. The tive slot and releases said resilient members for engage­
attachment provided takes up no more room than the ment with said socket member, and spring means between
casing itself and both parts which ultimately engage 15 the tube and the cylindrical casing in a direction pressing
each other are of such construction that rough usc and the wall of the respective enlargement snugly against the
mishandling wiII not damage their effectiveness. The configuration therein.
arrangement furthermore is such that an interior space is 3. A tube and shield assembly comprising a base hav­
left within the casing surrounding the glass envelope for ing a flat chassis contacting portion, laterally extending
any one of a number of related mechanisms or attach­ 20 ears having attachment means thereon, an upstanding
ments which might be employed with the casing. substantially cylindrical socket adapted for the reception
While I have herein shown and described my invention of an electronic tube, a tube in said socket and a rounded
in what I have conceived to be the most practical and protruding configuration on the side wall of said socket,
preferred embodiment, it is recognized that departures a one part substantially cylindrical casing of drawn sheet
may be made therefrom within the scope of my invention, 25 metal adapted to extend around and above the electronic
which is not to be limited to the details disclosed herein tube at a distance laterally spaced therefrom, said casing
but is to be accorded the full scope of the claims so as having a socket-engaging open end adapted to fit over
to embrace any and all equivalent devices. said socket with a snug sliding fit, said open end having
Having described my invention, what I claim as new a plurality of relatively wide circumferentially spaced

and desire to secure by Letters Patent is: 30 slots extending inwardly from the end and forming a
1. A tube shield mounting comprising a base having corresponding number of elongated resilient members,
a chassis contacting portion and attachment means said slots having progressively expanding wide entrance
thereon, a substantially cylindrical socket member on portions at ends thereof adjacent the open end of the
the base, a substantially cylindrical casing member casing and having enlargements no greater in transverse
adapted to extend around and above an electronic tube, 35 dimension than the transverse dimension of said con­
said casing member having a socket-engaging open end figuration intermediate outer and inner ends and at a
adapted to fit about said socket member with a snug location more nearly adjacent the outer ends whereby
sliding fit, one of said members having a protruding con­ when said casing is forced over the socket engagement
figuration on the side wall thereof, the other of said of the walls of one of said slots with said configuration
members having a plurality of circumferentially spaced 40 is adapted to spread adjacent resilient members of the
open slots extending inwardly from an end edge, said casing and guide said casing to a position wherein said
slots having portions at ends thereof adjacent the open configuration projects into the enlargement of the respec­
end of the member spreading progressively outwardly tive slot and said resilient members are released for
toward the end edge and having enlargements no greater spring-pressed engagement with said socket, and spring
in transverse dimension than the corresponding dimension 45 means between the tube and the cylindrical casing
of said configuration intermediate outer and inner ends biased against the casing in a direction pressing the wall
whereby when one said member is forced over the other of the enlargement snugly into engagement with the
said member engagement of one of said slots with said configuration.
configuration is adapted to guide said members to a posi­
tion wherein said configuration projects into engagement 50 References Cited in the file of this patent
with the wall of the enlargement of the respective slot.
UNITED STATES PATENTS
2. A tube and shield assembly comprising a base having
a chassis contacting portion, and attachment means 2,037,620 Dewhurst et at. ________ Apr. 14, 1936
thereon, a substantially cylindrical socket adapted for 2,453,172 Wilkie ________________ Nov. 9, 1948
the reception of an electronic tube and a tube in said 55
FOREIGN PATENTS
socket, a protruding configuration on the side wall of
said socket, a substantially cylindrical casing adapted 455,325 Canada ______________ Mar. 22, 1949
to extend around and above the electronic tube, said 622,075 Great Britain __________ Apr. 26, 1949
Aug. 11, 1959 L. R. WOODS 2,899,670
TUBE SHIELD ATTACHMENT

Filed Feb. 6, 1956

20
21'

.L-L,eOV.Ji!h1LPH WOOL)S
INVENTOR.
/h�t5/7e-,: ..6'eeAlt!'-': Worr�/,/ #�z'f'

V��AlG3 Q c� 'tA-

I"9rTO-€/vE�..s:
♦ Verso Filler Page ♦

United States Patent Office
1 2
2,935,549
TUBE SHIELD LINER
Leroy R. Woods, WbJttler, Callf., asslgnor to Inter­
national Electronic Research Corporation, Bnrbank,
Callf., a corporation of California
Application September 17, 1�56. Serial No. 610,187
5 Claims. (Cl. 174-35)
The invention relates to shields for electronic com­
ponents and has particular reference to a tube shield liner
and an appropriate shield which is particularly well
adapted to dissipating heat generated in the tube and
simultaneously minimizing the effect of vibrations of high
frequency which might be present in the chassis and be
otherwise transferred to the electronic component
through the shield.
Within the past few years attention has been directed
for the first time to improving tube shield assemblies with
special emphasis upon an improved transfer of heat gen­
erated during operation of the tube thereby to extend the
life of electronic components in the form of tubes with
glass envelopes. More stringent requirements brought
about by a greatly accelerated employment of electronic
equipment in jet aircraft and missiles have furthermore
made obsolete to a large degree tube shields previously
employed because of their inability to prevent vibrations
and especially high frequency vibrations from passing to
the rather delicate interiors of electronic tubes, thereby
causing early breakdown of these parts. Failure of elec­
tronic tubes in critical equipment is such that the likeli­
hood cannot be tolerated where it is at all possible to
g�u!rd against it and the expedients resorted to in order to
assure continued perfonnance have often resulted in
wholesale replacement of good tubes where they may
hn.ve been in use for a period of time.
Although some expedients have already been worked
out in the interest of supplying some of the needs herein
referred to, some of these innovations have been to a de­
gree complicated, thereby impairing widespread use be­
cause of price. On one or two other occasions where in­
expensive design has been attempted,· efficiency has been
cut down to a degree where such inexpensive variations
have only a very limited application.
It is therefore among the objects of the invention to
provide a new and improved liner for an electronic com­
ponent shield which is extremely simple, inexpensive to
manufacture and install and which at the same time is
particularly efficient and effective both with respect to the
transfer of heat and with respect to minimizing the effect
of high frequency vibrations on the electronic component.
Another object of the invention is to provide a new and
improved tube shield assembly which is so designed that
it can be constructed of raw materials in inexpensive form
and one capable in combination of virtually maximum
effectiveness by use of relatively a minimum number of
individual parts.
Still another object of the invention is to provide a new
and improved liner for tube shields which can be stamped
from a single piece of sheet material and which has such
form that it can be readily rolled into cylindrical shape
for easy insertion in the sleeve portion of a tube shield.
Still further among the objects of the invention is to
provide a new and improved tube shield liner material of 'TO bases 33 of fingers 31 are on the opposite side. By the
such form, construction and configuration that large
sheets .can be formed with a. single oper�ti�n in. �uc4
2,935,549
Patented May 3, 1960
fashion that they can subsequently be recut to various
sizes, thereby to accommodate tube shields of various
sizes without the necessity of maintaining a separate in­
ventory for each and every separate size.
IS With these and other objects in view, the invention
consists in the construction, arrangement and combina­
tion of the various parts of the device whereby the objects
contemplated are attained,as hereinafter set forth, pointed
out in the appended claims and illustrated in the accom·
10 panying drawings.
In the drawings:
Figure 1 is a perspective view of the tube shield as­
sembly incorporating the liner herein made reference to.
Figure 2 is a longitudinal sectional view on the line
15 2-2 of Figure 1 showing the tube shield combination in
place around an electronic tube.
Figure 3 is a plan view of a piece of sheet material
formed in a large piece with configurations arranged in
such manner that the large piece can be cut to a proper
20 size to fit some selected tube shield combination.
This is a continuation-in-part of my co-pending appli­
cation Serial No. 345,597, filed March 30, 1953, now
Patent No. 2,807,659.
In an embodiment of the invention chosen for the pur-
25 pose of illustration there is shown a tube shield combina­
tion consisting of a base 10, a sleeve 11, and a liner indi­
cated generally by the reference character 12. For pur­
poses of illustration the base is shown mounted on a
chassis 13 by means of screws 14. A socket 15 receptive
30 of a tube constituting a glass envelope or bulb 16 is shown
mounted within the base.
More particularly the base consists of a tubular ele­
ment having an inwardly bent flange 17 at the bottom
through which the screws 14 may pass.
35 The sleeve 11 is likewise tubular and has an annular
flange 18 at the top inwardly bent and relatively narrow
in width. For convenience in handling the sleeve a bale
19 is provided which engages the liner immediately
beneath the flange 18. In order that the sleeve may have
40 a snug sliding fit within the base, the outside diameter of
the sleeve is approximately equal to the inside diameter of
the base 10 but slots 20 are cut inwardly from the bottom
end of the sleeve throughout most of the height and
thereby incorporate into the sleeve a springy effect which
45 permits the sleeve to slide into the base and be main­
tained therein.
The liner 12 may be constructed in a large piece such
as that illustrated in Figure 3 or, should conditions war-
50 rant, may be constructed initially with only a length 21
and width 22 such as a single liner would have. As
shown the liner may be of virtually any length and even
in rolled stock if need be. While more conveniently the
liner will have a predetermined width 22 suitable to tube
5 shields of a given length, occasions may arise where the
5 piece of liner material may have a greater width and be
subject to being slit longitudinally along a line 21', there­
by to give it a desired width to fit a predetermined length
of tube shield sleeve, The inside circumference of the
60 sleeve will determine the length 21 and the location of
lines 22' and 23. This cutting along the lines 22' and 23
need be only through the metal portion separating rows
25,26,27, etc., of fingers 30,31.
The liner 12 as previously indicated is a single sheet of
6 spring-like material preferably metal chosen for its effec-
5 tive heat-conducting characteristic. The fingers 30 and
31 are identical but are reversed in position alternately
along each respective row; that is to say, bases 32 of
fingers 30 are on the same side of the row 2S whereas
same token apexes 34 of fingers 30 are on the opposite
. side of the row from apexes 35 of fingers 31. The apexes
 fa,9Sts,tJ49
3 4
and diagonal sides 36 and 37 in each instance are cut
clear.
The cut-out ·fingers 30 and 31 are pressed away from
the plane of the liner to the positions illustrated best in
Figure 2. As there shown the apexes 34 and 35 have 5
a somewhat fiat tip or end curved slightly so as to engage
over some area of the envelope 16 at each point of
contact.
It will be noted further that the rows 25 and 26 are
separated by an imperforate strip 40. Other imoerforate ] 0
strips 41 and 42 separate the other rows. /. Cla;!l ""�;'.!
43 appears on the lower side as viewed in Figure 3.
whereas a band 44 on the upper side is provided with a
specially shaped edge 45.
Assuming that the sleeve 11 takes a liner or insert 15
having a length 21 and breadth 22, a liner cut (0 that
length is rolled into a cylindrical shape with the rows
25, 26, 27, etc., taking an annular shape and with the
fingers pointed inwardly. The rolling is made smaller
than the inside circumference of the liner with edg.::s
overlapping and the liner inserted into the sleeve from
the bottom of the sleeve, as viewed in Figure 2. The
liner is pushed into the sleeve until the edge 45 strikes
the flange is, ut which time the overlapped edges
of the liner are released. The length of the liner
is cut to such dimension that when released the edges
will approximately abut each other. No other fasten­
ing means need be provided inasmuch as the spring­
like expansion of the liner itself coupled with the friction­
al engagement between the edges of the liner and the in­
side wall of the sleeve will be sufficient to hold the liner
in proper position.
With the liner inserted in the sleeve, the sleeve is then
pressed down over the glass envelope 16 within the base
and the tendency of that application will be sufficient to
preS5 the liner frictionally upwardly to its uppermost
position and to bind the sleeve within the base.
There has accordingly been herein shown and described
a tube shield liner of very simple construction, namely,
consisting of a single sheet of material stamped, however,
in such form that a multiplicity of spring fingers extend
outwardly from the plane surface of the sheet. The
fingers are regularly and uniformly spaced and alternately
positioned so as to maintain a maximum of structural
rigidity and at the same time afford a maximum amount
of springy consistency to the apexes where they are uni­
formly in engagement with the glass envelope through­
out its exterior. It will be noted further that although
there are openings in a direction laterally or edgewise
of the diagonal sides of the individual fingers, neverthe­
less there is a substantially imperforate presentation of
the liner as a whole to the glass envelope. Hence, al­
though there is ample space for ventilation of heat around
the tube shield, nevertheless there is a maximum amount
of electrical shielding effect for the tube. The multi­
plicity of area contacts at the apexes provides for effective
removal of heat by conduction and by the same structure
passage 01 vibrations of high frequency from the chassis
through the base to the tube envelope is minimized. The
composite structure made as shown and described of sim­
ple parts and simply constructed is effective in the pro­
duction of an effective combination with a minimum of
operations and amount of material.
While I have herein shown and described my inven­
tion in what I have conceived to be the most practicr.l
and preferred embodiment, it is recognized that depar­
tures may be made therefrom within the scope of my
invention, which is not to be limited to the details dis­
closed herein but is to be accorded the full scope of the
claims so as to embrace any and all equivalent devices.
Having described my invention, what I claim as new
and desire ot secure by Letters Patent is:
1. A vibration dampening and heat conducting liner
for an electronic component shield of hollow cylindrical
form comprising a rectangular piece of sheet metal having.
a width substantially equal to the length of the shield
and a length substantially equal to the inside circumfer­
ence of the shield, and a pattern of substantially uni­
formly spaced rows of tapered spring fingers stamped
from said sheet metal and spaced equidistant one finger
from another along each TOW, and an imperforate straight
flat band separating each row, each finger having a
trapezoidal form with the base attached to the sheet and
a portion thereof adjacent the apex extending outwardly
in a position substantially parallel to said flat bands and
having an area curved to form said portion into a flat
component contacting area, said fingers in each row ex­
tending alternately in opposite directions and on the same
face of said piece of sheet metal whereby to space said
contacting areas at alternately opposite locations and
adjacent fingers of adjacent rows facing in opposite direc­
tions whereby to balance the co n tactin g areas about the
liner.
2. A vibration dampening and heat conducting lining
20 material for elec:tronic tube shields comprising a piece of
sheet metal of length and breadth exceeding the length
and inside diameter of a tube shield, a plurality of equally
spaced parallel rows of spring fingers extending in one
direction of said piece, each spring finger having a rela-
25 tively wide base attached to the piece of sheet metal
and a relatively narrow apex raised a distance equal to a
fraction of its length from the surface of said piece of
sheet metal and substantially covering a hole from which
the finger is stamped, each said apex having an aNa sub-
30 stantially conforming to an adjacent area of said tube,
said bases of adjacent fingers of each row being located
alternately on opposite sides of the row and on the same
face of said piece of sheet metal, said piece ha'/ing por­
tions thereof parallel to and between the rows adapied
35 to be cut in one direction and having portions thereof
,transverse to the rows between fingers adapted to be cut
whereby to enable a plurality of liners for individual
tube shields to be cut from a single piece of preformed
sheet metal.
40 3. A vibration dampening and heat conducting tube
shield assembly for the glass envelope of an electronic
tube comprising an imperforate hollow base, a hollow
sleeve having an outside diameter substantially equal to
the inside diameter of the base whereby to provide a
45 spring fit for said sleeve in said base, a liner for said
sleeve comprising a single piece of sheet metal having a
width substantially equal to the length of the sieeve and
a length substantially equal to the inside perimeter of the
sleeve whereby end edges of said liner abut when in
50 place; and a series of spaced parallel perimetral rows of
spring fingers stamped from said sheet metal, each finger
having a relatively wide base attached to the piece of
sheet metal and a relatively narrow apex removed from
the plane of said piece of sheet metal, said fingers having
55 positions substantially covering the holes frem which they
are cut and providing laterally open spaces for circula­
tion of air between the exterior surface of the liner and
the envelope, the fingers in each row having bases on
alternately opposite sides of the row, each finger having
60 an area at the apex bent to a position substantially parallel
to said sleeve and being curved to conform to the shape
of the envelope whereby to provide a uniform pattern
of spring fingers encompassing and covering said en­
velope and pressing uniformly inwardly against the en-
65 velope and outwardly against the sleeve.
4. A vibration dampening and heat conducting tube
shield assembly for the glass envelope of an electronic
tube comprising an imperforate hollow cylindrical base,
a hollow cylindrical sleeve having an outside diameter
70 substantially equal to the inside diameter of the base and
having deep slits extending inwardly from an open end
whereby to provide a spring fit for said sleeve in said
base, and a bale at the opposite end of said sleeve, a
liner for· said sleeve comprising a single piece of sheet
'15 metal having-a width 'substantially equal to the combined
 g,9S�,649
5 6
length of the slit and unslit portions of the sleeve and a throughout a portion of their perimeters from the sheet
length substantially equal to the inside circumference of and spaced one finger from another along each row,
the sleeve whereby end edg�s of said liner abut when in each finger having a base attached to the sheet and the
place, and a series of equally spaced parallel annular rows apex free of the sheet extending outwardly of the plane
of spring fingers stamped from said piece of sheet metal I of the sheet on the same side thereof, each said finger
and located on the same face of said piece of sheet metal, having an outermost portion free of the sheet and bent
each finger having a relatively wide base joined to the to a position and form shaped in conformance with the
piece of sheet metal and a relatively narrow apex, the form of an adjacent portion of said component, each
fingers in each row having bases on alternately opposite said finger presenting a component contacting area
sides of the row and the bases of fingers of adjacent rows 10 initially parallel to the sheet, said sheet being substantially
bc-jng opposite each other, each finger having an aWl imperforate except for spaces opening laterally between
at the apex bent to a position substantially parallel to the plane of said sheet and the portions of the sheet form­
the face of the envelope and being curved to conform to ing said fingers whereby to form a continuous electrically
the face of the envelope whereby to provide a uniform shielding sheet, said sheet being adapted to be cut between
pattern of spring fingers pressing simultaneously inwardly 11 rows of said fingers in one direction and between fingers
against the envelope and outwardly toward the liner and in a transverse direction whereby to enable a plurality of
occupying an area encompassing and covering said en­ liners for individual ;tube shields to be cut from a single
velope. piece.
5. A device for use with an electronic component
having an exterior surface of predetermined form com- 20 References Cited in the file of this patent
prising a vibration dampening and heat conducting liner
UNITED STATES PATENTS
for an electronic component shield of hollow elongated
form said liner comprising a sheet of heat conducting and 514,313 Eckstein _______________ Feb. 6, 1894
electric shielding material having a width exceeding the 936,183 Stein ___________________ Octo 5, 1909
length of the shield and a length exceeding the inside 15 1,548,809 Jones _________________ Aug. 4, 1925
perimeter of the shield, a pattern of substantially uni­ 1,839,284 White __________________ Jan. 5, 1932
formly spaced rows of tapered spring fingers stamped 2,745,895 Lideen _________________ May 15, 1956
May 3,1960 L. R. WOODS 2,935,549
TUBE SHIELD LINER

Filed Sept. 17, 1955


United States Patent Office
1 2
2,951,103
TUBE SHIELD AND BASE ASSEMBLY
Leroy Ralph Woods, Whittler, CaJif., assignor to Interna­
tional Electronic Research Corporation, Los Angeles,
Calif., a corporation of California
Flled Apr. 23, 1956, Ser. No. 580,015
1 Claim. (Cl. 174-35)
The invention relates to tube shields and has special 1 5
reference to parts constituting a base and a shield for
attachment ,thereto under circumstances where the shield
or envelope portion of the assembly may be specially
equipped to hold the glass bulb of the tube in some speci­
fied manner.
After it was discovered that for certain purposes an
advantage exists in enclosing electronic tubes in a metallic
shielding envelope for a long time the shielding consti­
tuted no more than use of a cylindrical sheet metal en­
velope pressed .into position around the tube and socket 25
and somewhat loosely attached to an anchoring means
on the chassis. - Recent developments, however, have
emphasized the advantages and necessity for not only
shielding the tube against the passage of radio waves -of
certain frequencies but also, and more especially, for
the purpose of conducting heat away from the glass bulb
without inducing temperature cracks therein and also for
improving the mounting of the electronic ,tube, thereby
to minimize deteriorating effects which may result from
use of the tubes in locations where sustained high fre­
quency vibration effects might be present.
Construction of tube shields or tube shield envelopes
to satisfy needs of this nature has been found incomplete
where resort might be had to bases of conventional de­
sign for retention of the shielding envelope. Unless the
connection between the base and the envelope has a heat­
conducting character comparable to the heat-conducting
faculties of the shield assembly, many of ,the advantages
of the latter are nullified. Also, if the shielding envelope
is to be effective in minimizing the effects of high fre­
quency vibrations, the connection between the tube shield­
ing envelope and the base must be one capable of main­
taining a sustained snug connection under such circum­
stances.
It .is therefore among the objects of the invention to 50 cylindrical collar extends upwardly from the flange and
provide a new and improved tube shield and base assem­
bly which is simple in construction and which at the same
time is positive in its attachment, thereby to satisfactorily each .includes an inwardly projecting lip 29 tilted at an
conduct heat ,through the junction. 55 oblique angle, as shown to good advantage in Figure 2.
Another object of the jnvention is to provide a new
and improved tube shield and base assembly which pro­
vides a firm grip of one with respect to the other and
which is sufficiently rugged and permanent to maintain its
action under conditions where high frequency vibrations
60
might exist.
Still another object of the invention is to provide a
new and improved tube shield base which is sufficiently
simple as to permit fabrication from sheet metal and
which is so formed that it is rigid, durable, and capable 65 17 at its lower end substantially equal to the inside di-
of securing and maintaining a positive grip upon a tube
shield envelope mounted thereon.
Still further among the objects of the invention is to
provide a base for a tube shield envelope and an envelope
for connection therewith, both of which are relatively
simple and inexpensive sheet metal parts so constructed 70 inside wall of the collar. When the envelope is appiied
that they can be quickly engaged one with respect to the
2,951,108
Patented Aug. 30, 1960
other, thereafter 10 maintain the engagement but which
when necessary can be easily disengaged for facilitating
removal and replacement of the tube within the shield.
With these and other objects in view, the invention
5 consists in the construction, arrangement and combina­
'iion of the various parts of the device whereby the ob­
jects contemplated are attained, as hereinafter set forth,
pointed out in the appended claim and illustrated in the
'accompanying drawings.
10 The invention herein disclosed and claimed is a con-
tinuation-in-part of co-pending application Serial No.
345,597, filed March 30, 1953, now Patent No. 2,807,-
659.
In ,the drawings:
Figure 1 is a longitudinal sectional view of the tube
shield and base assembly shown in position around a con­
ventional electronic tube and socket therefor.
Figure 2 is an exploded perspective view of the tube
shield envelope and base in the relative positions they
20 would have prior to engagement.
Figure 3 is a perspective exploded view of a second
form of the device.
Figure 4 is a fragmentary elevational view of connect­
ing elements of the base and shield of Figure 3 shown
prior to connection.
Figure 5 is a view similar to Figure 4 but showing the
elements connected.
Figure 6 is a longitudinal sectional view taken on the
line 6-6 of Figure 5.
30 In an embodiment of the invention chosen for the pur-
pose of illustration there is shown a chassis 10 mounting
a retainer 11 to which a socket 12 is attached. An
electronic tube is mounted in the socket and a glass bulb
13 of the tube extends upwardly out of the socket.
35 The tube shield and base assembly comprises in the
main two separate parts, namely, a base indicated gen­
erally by the reference character 16 and a tube shield
device indicated generally by ,the reference character 17.
40 For purposes of illustration the tube shield is illustrated
as one containing a liner having a plurality of spring-like
contact elements 18 which create a spring contact be­
tween the bulb 13 and the tube shield 17. The tube
shield proper may appropriately be described as a cylin­
drical envelope. A flange 19 at the top of the envelope
45 has an overhanging portion adapted to retain beneath it
a spring 20 for pressing upon the outer end of the bulb.
The base 16 comprises an outwardly extending flange 21
having holes 22 through which screws or rivets 23 may
be extended to secure the base to the chassis 10. A
at the top of the collar is a series of four tabs 25, 26, 27
and 28. The tabs are identical in each instance and
The tube shield or envelope 17 has four correspond­
ing base gripping portions 30, 31, etc., each portion in
turn including a laterally open, obliquely disposed slot
32 matching jn obliquity the lips 29. A short finger ele-
ment 33 underlies the slot and a long finger element 34
lies over the slot. Tab-receiving spaces 35 are pro­
vided for by the short finger element having a width
more than sufficient to receive any one of the tabs 25,
26, etc. By having the diameter of ,the shield or envelope
ameter of the collar 24, or in fact by having the base­
gripping portions extended very slightly, they can be
sprung inwardly when the envelope is pressed downward­
ly inside of the collar and the outer surfaces of the
envelope will therefore be in snug engagement with the
to the base, the base-gripping portions are pushed within
 2,951,108
3 4
the collar and the spaces 35 extended over the tabs and
the inwardly extending lips 29. The envelope is then
rotated slightly less than 90 degrees and the lips slide
down the oblique slots 32 to Ll}e bottoms of the slots
where they are maintained partly by the expanding re­
silience of -the lower end of the envelope and in part by
the action of spring 26 which tends to force the envelope
outwardly relative to the base. Windows 36 assist in
passing heat by radiation from the glass bulb to the ex­
terior.
In a second form of the invention illustrated in Fig­
ures 3, 4, 5 and 6 there is shown an envelope 40 which
may consist of a single metal stamping continuous
throughout its perimeter, on the upper end of which is a
flanged ring 41. Windows 42 may be provided if needed
for facilitating the escape of heat through the envelope.
At a base end of the envelope is a plurality of base­
gripping portions 43, on one side of which is a slot 44,
below which is a short finger element 45, and above
which is a long finger element 46. A base indicated gen­
erally by the reference character 47 has a flange 48
adapted to be secured to a chassis 49 by means of screws
or rivets 50. A collar 51 extends upwardly of the flange
48 and at the upper rim of the collar are inwardly extend­
ing tabs 52, each tilted at an oblique angle corresponding
to the obliquity of the slots 44. Adjacent open ends of
the slots are spaces 55 materially wider at the narrow
end than the length of the tabs 52 and being progressively
wider toward the open end. Sloping sides 56 and 57
assist in guiding the envelope to its position with re-
spect to the tabs 52. After the envelope has been pushed
downwardly until the tabs engage bottoms 58 of the
spaces 55, the envelope is rotated, and, the tabs being
at the same angular pitch in the slots, a snug engage-
enough to withstand disturbance by the presence of high
frequency vibrations.
While I have herein shown and described my inven­
tion in what I have conceived to be the most practical
5 and preferred embodiment, it is recognized that depar­
tures may be made therefrom within the scope of my
invention, which is not to be limited to the details dis­
closed herein blJJt is to be accorded the full scope of the
claim so as to embrace any and all equivalent devices.
10 Having described my invention, what I claim as new
and desire to secure by Letters Patent is:
A tube shield assembly for application over an elec­
tronic tube comprising; a substantially cylindrical base
adapted to be mounted in an upright position on a chassis,
15 said base having a plurality of circumferentially spaced
intumed tabs at its upper edge, a shielding tube and
retaining envelope comprising a sheet metal cylinder hav­
ing an inturned flange at its upper end, a spring inside
said shielding tube, bearing against said flange and
20 adapted to press against an electronic tube to urge said
shielding tube upwardly, the lower end of said shielding
tube being of a diameter to be snugly received within
said base and being provided with a plurality of down­
wardly open notches corresponding to said tabs on said
25 base and of a size to readily receive the same, a slot ex­
tending helically from the uppermost portion of each
notch and adapted 10 receive a corresponding tab, each
of said slots extending helically downwardly toward but
short of the lower end of said shielding tube.
30
References Cited in the file of this patent
UNITED STATES PATENTS
337,867 Richards Mar. 16, 1886
831,224
______________

ment is readily accomplished. 35 Hammer Sept. 18, 1906

2,367,458
_______________

There has accordingly been disclosed herein a tube Coplen Jan. 16, 1945
2,646,460
________________

shield envelope and base· assembly which is sufficiently Del Camp ______________ July 21, 1953
simple in design to be capable of being fabricated by
stamping from light gage sheet metal and which by rea- FOREIGN PATENTS
40
son of the relationship of the parts provides an easily con- 11,284 Great Britain May 25, 1908
of 1908
__________

nectable firm grip between the envelope and base tight

enough for the transfer of heat by conduction and secure 649,059 Great Britain ____ ______ Jan. 17, 1951
 Aug. 30, 1960 L. R. WOODS 2,951,108
TUBE SHIELD AND BASE ASSEMBLY

Filed April 23, 1956

J6
JO ���====��E

19

i /"'/7
I
I
:5'6 --I---tt""-
;1
I
i
I

1
'I
11

:1
�E.eOV �1'9��# 1f/OOL)S
INVENTOR.
#Ve.tf/le� .tf>eeA/e� Wor/'e/�#e/'z&.

U��,--
RrrtJeNEY'S.
♦ Verso Filler Page ♦
United States Patent Office
1 2
%,973,400
HEAT TRANSFER LINER AND TUBE SHIELD
JohnC. McAdam, Burbank,Calif., assigoor to Inter­
national Electronic Research Corporation, Burballk,
Calif., a corporation ofCalifornia
Filed Aug. 1, 1958, Ser. No. 752,457
3Claims. (Cl. 174-35)
The invention relates to tube shields for use with
electronic components and electronic tubes wherein the
tube shield is so constructed that it is capable of pick-
ing up heat from the surface of the component or tube
and transferring it through a surrounding tube shield
envelope to the base primarily by conduction, at which
point heat is dissipated to the mounting.
The invention further comprehends a heat dissipating
tube shield which is also constructed in such fashion that
the effects of vibration in the mount which might be
detrimental to the performance or long life of the com-
ponent are substantially minimized.
In relatively recent years the industry has learned of
the marked advantages in carrying away from electronic
components and tubes heat which is often generated in
substantial quantities when the tubes are performing in
their normal fashion. As the conditions have changed
to circumstances wherein more and more tubes are com­
pacted within a limited space, the heat dissipating prob­
lem has correspondingly increased. Also for the same
reason .where many tubes and components are com­
pacted within a small space on vehicles, for example,
which suffer an exceptional amount of high frequency
vibration, the .mortality in electronic tubes has increased
appreciably and to the extent that positive means for
minimizing the effect of vibration as well as heating has
become highly desirable.
Some tube shield devices have been developed to
satisfy the requirements mentioned and performance­
wise these devices have been acceptable and effective.
Due to the fact, however, that the tube shield liners are
subjected to repeated heating and cooling cycles, some
problems have arisen with respect to the form and char­
acter of the. resilient tube shield liner in order that it
continue its effective performance over long periods of
time without the necessity of frequent servicing. Many
of the tube shield liners heretofore employed have either
·been unnecessarily costly for a great variety of installa­
tions where dependability has been a primary requisite
or have limited effectiveness where they have been con­
structed inexpensively and in a single piece due to cer­
tain limitations in the mode of manufacture.
It is therefore among the objects of the invention
to provide a new and improved heat transfer liner and
tube shield combination wherein the elements are par­
ticularly simple and non·complicated with respect to
manufacturing requirements and technique and which at
the same time incorporate many of the advantages here­
tofore available only in the more intricate heat dissipat­
ing tube shields and liners.
Another object of the invention is to provide a new
and improved heat transfer liner for electronic tube
shields which can be constructed of a single strip of re­
'silient material in such fashion that a very high degree
of resilient contact is provided on the inside capable of
engaging a substantial portion of the tube or component
and wherein a lesser portion on the outer side is made
,use of for, engagement with the. surrounding sleeve, theI:�-
2,973,400
Patented Feb. 28, 1961
by making most effective use of the available area of
the liner.
Still another object of the invention is to provide a
new and improved heat transfer liner for electronic tube
5 shields which can be constructed by forming a strip of
selected material essentially flat in an initial form such
that after the strip is curved and fitted within a sleeve,
the lin�r assumes a new and different form which is
highly advantageous from the point of view of contact,
10 heat dissipation and the inhibiting of the transfer of vi­
brations.
Still further among the objects of the invention is to
provide a new and improved heat transfer liner and
tube shield combination wherein the parts cooperate
15 with each other in such fashion that the respective en­
gagements of the liner with the tube and the liner with
the surrounding sleeve are proportioned in order to have
the combination perform in the most effective manner.
With these and other objects in view, the invention
20 consists of the construction, arrangement and combina­
tion of the various parts of the device whereby the ob­
jects contemplated are attained, as hereinafter set forth,
pointed out in the appended claims, and illustrated in
the accompanying drawings.
25 In the drawings:
Figure 1 is a side perspective view of the heat trans­
fer liner and tube shield combination shown mounted
upon a base in position of engagement with an elec­
tronic tube.
30 Figure 2 is a plan view of the liner in the form had
prior to insertion within the sleeve.
Figure 3 is a plan view of the liner and sleeve com­
bination.
Figure 4 is an end elevational view of a portion of
35 the liner in the form had prior to insertion within the
sleeve.
In an embodiment of the invention chosen for the
purpose of illustration there is shown a base 10 adapted
for attachment to some conventional chassis by means
40 of holes 11. The base is so constructed that it is adapted
to receive an electronic component in the form of an
electronic tube indicated generally by the reference char­
acter 12, the tube being provided with a base portion 13.
Fitting slidably over the base is a metallic sleeve 14
'6 which makes use of slits 15 having apertures 16 therein
for reception of dimples 17 on the exterior of the base
whereby to firmly but releasably anchor the sleeve to
the base. At the upper end of the sleeve is an annular
inwardly turned flange 18 defining a relatively large
60 opening 19.
A liner is indicated generally by the reference char­
acter 20 and is shown in position between the tube 12
and sleeve 14 in Figure 1. The liner is shown separate·
ly in Figures 2 and 4 in the form existing prior to in·
115 sertion of the liner within the sleeve.
More particularly, the liner comprises a single sheet
of metallic material which has a resilient character de·
pendable under all conditions of use and which will main­
tain its resilient character with a high degree of efficiency
60 when subjected to frequent heating and cooling cycles
over long periods of time. Beryllium copper has beeD
found to have the desired characteristics and reference is
made to this material by way of example.
When the liner is formed initially, there are provided
65 a series of spaced parallel ribs 21 which in plan view have
the form of elongated rectangles and whose outer surfaces
22 are substantially flat. In the chosen embodiment the
ribs are uniformly and equally spaced one from another,
the space between ribs being substantially greater than
70 the breadth of the ribs. Between the ribs are intermediate
portions 23 substantially wider than the ribs and entirely
1illing .th� �paces. �tween. the ribs.. The intermediate
 �,97a;400
3 4
portions have arcuate sides 24 which join longitudinal the component so that there will be virtually no uneven­
side edges of the ribs with the intermediate portions. The ness in the pick up of heat or in the absorption of vi­
arcuate portions are gently arcuate at the point of junction brations.
or tangency with the intermediate portion but joins the While I have herein shown and described my invention
ribs at a relatively sharp angle. The ribs and intermedi- 5 in what I have conceived to be the most practical and
ate portions may be formed in an acceptable manner preferred embodiment, it is recognized that departures
either by rolling or die forming. It will also be appre­ may be made therefrom within the scope of my invention,
ciated that the strips can be of any convenient length which is not to be limited to the details disclosed herein
longer than might be needed for any single tube shield but is to be accorded the full scope of the claims so as to
liner and that the strip can be cut at will anywhere along 10 embrace any and all equivalent devices.
its length. Having described my invention, what I claim as new
As shown in Figure 3,. cuts are made adjacent the ribs and desire to secure by Letters Patent is:
through the arcuate sides leaving short legs in order to 1. A heat dissipating and vibration resistant shielding
have the ribs overlap each other, thereby to anchor the liner for use in holding a circumferentially arcuate elec­
ends with respect to each other. Due to the springiness 15 tronic component in a surrounding sleeve comprising a
of the material it is not necessary that the ribs interlock strip of homogeneous inherently resilient heat conducting
precisely and hence the endmost ribs can be retained metallic material, said strip comprising pairs of outwardly
precisely the same breadth as all of the other ribs. extending corners defining a series of spaced ribs each
After the liner has been formed in the manner de­ being of predetermined breadth between the corners and
scribed, it is turned upon itself as illustrated in Figure 3 20 presenting areas adapted to engage the surrounding sleeve,
and inserted within the sleeve 14. The number of sec­ said liner including intermediate portions of said metallic
tions employed should be such that the liner is not material between corners of adjacent ribs having a breadth
stretched unnecessarily when placed within the sleeve but greater than the breadth of the ribs, each said portion
rather should be such that if it is necessary to help the having side sections of arcuate form curving towards each
liner fit within the sleeve, it can be puckered slightly. 25 other and joining a mid-section therebetween with a tan­
Due to the inherent resiliency in the intermediate por­ gential junction, whereby when the liner and sleeve are
tions, they will assimilate a very substantial degree of applied to a component the intermediate portions through­
adjustment. out the span thereof between adjacent corners yield resili­
When the combined liner and sleeve is to be employed ently upon engagement with the component.
in the manner shown in Figure 1. they are slid together 30 2. A heat dissipating and vibration resistant shielding
over the outer end of the tube during which operation liner for use in holding a circumferentially arcuate elec­
the intermediate portions 23 will yield substantially, there- tronic component in a surrounding sleeve comprising a
by to accommodate the surface of the tube. The yielding continuous strip of homogeneous inherently resilient heat
is sufficiently variable so that even though the outside di­ conducting metallic material having opposite side edges,
mension of the tube varies considerably throughout its 35 said strip comprising pairs of clearly defined outwardly
length, the intermediate portions will form an effective extending corners defining a series of spaced ribs each
.
engagement. being of fixed breadth between the corners and present­
Furthermore, by reason · of the same inherent resilience ing outwardly facing areas adapted to engage the sur­
throughout the entire area of the intermediate portions, rounding sleeve, said liner including intermediate portions
the liner can be used repeatedly for different tubes under 40 of said metallic material between corners of adjacent ribs
circumstances where the outside diameters of the tubes having a breadth greater than the breadth of the ribs, each
might vary considerably from one tube to another. The said portion prior to insertion of the component having
construction of the liner is such that it will readily ac­ side sections of arcuate form curving towards each other
commodate itself to a great number of variations in size and joining a mid-section therebetween with a tangential
and contour without suffering any permanent deforma- 45 junction, each said mid-section subsequent to insertion of
tion. a component having an outwardly arcuate shape and tan­
Under certain special circumstances as, for example, gential junctions with the adjacent arcuate sections, the
where a liner might need to be made of material some­ span of said outwardly arcuate shape and the lengths of
what heavier in gage than on other occasions, added ver­ the arcuate sections being determined throughout the
satility may be built into the liner by providing slots 27. 50 length of the (,;omponent by the perimeter of the compo­
The slots 27 have approximately the proportionate width nent whereby, when the liner and a sleeve are applied
shown and need have a length. no greater than the breadth to a component, the ribs at the side edges of the liner are
of the intermediate portions between the arcuate sides. interlocked at the outwardly facing areas and the area
An effective pattern is one illustrated in Figure 2 wherein of engagement of the liner with the component exceeds
the slots are parallel and uniformly spaced throughout 55 the area of engagement of the liner with the sleeve.
the length of the intermediate portions and wherein they 3. A heat dissipating and vibration resistant shield for
provide resilient patohes which are more or less square. use in holding a circumferentially arcuate electronic com­
When in use, whether with or without slots, the inter­ ponent comprising a substantially cylindrical sleeve and
mediate portions exert a resilient force which not only a liner therein of homogeneous inherently resilient metal­
impresses those portions into contact with the surface of 60 lic material, said liner comprising pairs of clearly defined
the component or tube, but also exerts a resilient pressure outwardly extending corners defining a series of spaced
outwardly, thereby pressing the surfaces of the ribs into ribs each being of predetermined breadth between the
firm heat conducting contact with the surrounding sleeve. corners and presenting outwardly facing areas adapted to
The firmness of contact at all areas is instrumental in
engage the surrounding sleeve, said liner including inter­
providing a . path for conducting heat generated in the 65
mediate portions of said metallic material. between cor­
tube through the liner, the sleeve and the base to the
ners of adjacent ribs having a breadth greater than the
chassis. It will be appreciated also that some heat is dis­
breadth of the ribs, each said portion prior to application
sipated by radiation from the exterior of the sleeve.
Moreover, the same resilient contact by which the inter­ to the component having side sections of arcuate form
mediate portions uniformly engage all variations in the 70 curving towards each other and joining a mid-section
surface of the tube is also instrumental in minimizing therebetween with a tangential junction, each said mid­
the transfer of damaging vibrations from the chassis to section subsequent to insertion of a component having an
the tube. A uniformly large area of contact is provided outwardly arcuate shape and tangential junctions with
which is effective in spreading the desirable effect uni­ the adjacent arcuate sections, the span of said outwardly
formly over substantially the entire exterior surface of 75 arcuate shape and the lengths of the arcuate sections being
 2,973,400
5 6
detennined throughout the length of the component by 2,807,659 Woods ________________ Sept. 24, 1957
the perimeter of the component whereby when the shield 2,905,742 Woods ________________ Sept. 22, 1959
is appli ed to the component the area of engagement of OlHER REFERENCES
the liner with the component exceeds the area of engage-
l5 Publication I, "Heat Reducing Tube Shield," Collins
nent of the liner with the sleeve.
Technical Bulletin 301 (4 pages), copy received U.S.
Patent Office Mar. 23, 1956.
References Cited in the file of this patent
Publication 11, "Electron Tube Reliability," published
UNITED STATES PATENTS in Electromechanical Design, January February 1958
-

2,7';5.895 Lideen ________________ May 15, 1956 10 (pages 16 and 17 relied on).
 Feb. 28, 1961 J. C. MCADAM 2,973,400
HEAT TRANSFER LINER AND TUBE SHIELD

Filed Aug. 1, 1958

II
fJ!I, 4, c2 et) 22
II
�i?4 �/ �
cl �g)
?.1 Z� c2

l'7
f
7
. 1
21
{
J
27
(
1 1 1 1 11'
I

1I1 1 1 I I 1 1 . 7\
I11
11 ( .

1 I1
 TUBE ENVELOPE TEMPERATURE
By WALTER R. laNES

Panel on Electron T1tbes, 346 Broadway, 8th Floor, New York l3, N.Y.

Operating temperature of tube envelopes becomes
increasingly important as the size of electronic
equipment is reduced. If the same amount of
power is to be dissipated within twO enclosures
one of which is much smaller than the other, the
temperature of the smaller enclosure will rise to
much higher values than that of the larger. This
fact is borne Out in Table I, which indicates both
areas and maximum dissipations for the bantam or
T9 tube, the miniature or T5! tube and subminia­
ture or T3 tube. It will be noted that as the sizes
decrease, the value of watts per square inch is
increase d and rhe maximum bulb temperature is
inHeased correspondingly. The l i fe of a vacuum
cube is materially affected by the operating tempem­
cure of its bulb, as well as its other pans.
TABLE I - Bulb 'Temperature at Sea Level
There are some absolure limits on the permissible
gl�ss temperature, one being the sGftening point of
the glass and another the point at which apprec)able
conductivity occurs - called electrolysis. Below
these limits there is an indefinite region in which
varying kinds a'tld degrees of trouble are encoun­
tere_, especially owing [0 evolution of gas from the
bulb itself., the getter and other: tube pans. The
temper·acure of concern is that at the honest spet
TABLE I-Bulb Temperature
at Sea Level
19 T5�
Bulb Area (Sq In.) 10.5 4.1
Mu. W.Hs Dissipation 18.7 16.8
Watts per Square Inc" 1.8 4.1
Mu. Bulb 1emp deg C 160 255
Ambienl 23C
reached while the tubes were being pumped during
production, it is likely that varying amounts of gases
will adversely affect tube life.
Normally, the function of the getter that produces
the silver-like deposit or the black deposit on some
of the newer tubes is to provide a means for
removing any gases that may subsequently be set
free during the operation of the tube. There is a
limited amount of gas that t h is getter mat(frial (all
safely piek up. Amounrs beyond this will result'
in the tube's gas content being materially increased.
(n addition. jf the glass bulb should be heated
sufficiently .[he getter patch may be caused [0
migrate or leave the bulb.
TABLE II - Bulb Temperatures 23° C
Ambient at Sea Level
It may redeposi t itself on some cooler part of the
tlIbe so that a considerable amount of gas trapped
by the getrer will now be released and may not
reCGmbin� when [he getrer condenses on the cooler
portions of the wbe. In this instance, then, the
gas coment would also be materially increased.
Should the getter condense on the mka supports
of dYe cube there is a possibiJ,ity that leakilge be­
tween elements supported in the mica may be in­
creased. This leakage may a ffect petformance
materially.
As seen frem Table I, l' tendency to decrease the
size of electronic: gear aggravates the bulb-tempera­
tUre condition. Tube. life, in gene ral, can be ex­
T3 tended by maintaining low temperatures for the
glass envelope. This is especi3oc1ly important in
high -p ower output tubes because of their higher
1.7
plate and cathode dissipations. The remperature
7.8
rise in rhe envelope may be limited by: reductidn
4.6 c;)£ tOtwi tube dissipation; provisien for improved
280 ventilation; maintenance of low ambient -tem­
peratures.

on the envelope, which usually occurs midway be­

TABLE II-Bulb Temperatures
tween the top and bottom micas. The location can 23° C Ambient at Sea Level
be readily found by the use of temperature sensitive
lacquers marketed by Tempil Corporation, 11 West
PerC2nt Maximum Plate Dissipation
25th Street, New York City.
During manufacture, while the vacuum tube is on Tvpe Bulb 20 40 60 80 100
the production pumping set-up, the envelope and
the metal parts within the tube structure are heated 12AU7 16Y2 77C 100C 118C 133( 146C
to much higher temperatures than those to which 6C4 T5'� 64 82 98 113 125
they would normally be subjected during operation 6AH6 15'4 88 103 116 126 132
in order that all of the absorbed and adsorbed 5U4G 5T16 105 116 127 138 149
gases may be removed. If, however, during ope­
5687 1631 123 140 155 155 183
ration, the operating temperature of the envelope
or the parts themselves exceeds the temperatures
Reprinted from "Electronics", by courtesy of the publishers.

Radiotronics 42 April, 1956

 In general, the envelope temperature of small
receiving-type power tubes should be kept below
175 deg. centigrade for increased reliability. The
chief effect of high temperature on vacuum tubes
is nOt a sudden change in operating characteristics
but a gradual deterioration of characteristics. Table
II indicates the operating bulb temperatures for five
types of tubes, having various sized envelopes for
plate dissipations ranging from 20 percent up to
maximum rated dissipations. This gives an idea of
the extent to which it is possible to reduce bulb
temperatures by decreasing the total tube plate
dissipation.
The ultimate bulb temperature depends not only
upon the dissipation within the tube itself but also
upon the temperature of the surrounding air im­
mediately adj acent to the tube envelope. Table III
shows how these ambient temperatures affect the
bulb temperature for various watts per square inch
dissipation. From these data it is apparent that
precautions must be taken to keep the ventilation
around the tubes such that the temperature will be
as low as possible.
Table III Sea Level Bulb Tem­
- "Data
peratures vs. Dissipations and
Ambient Temperature Variations.
Watt s/l n2 ====> 1 2 3 4
So far, sea-level altitudes have been assumed.
Many tubes operate at high altitudes some or all of
the time. This environment aggravates the cooling
problem still more since the density of the air de­
creases with altitude. The decreased effective cool·
ing of a tube at higher altitudes requires that the
tOtal tube dissipation be derated in order nOt to
exceed critical bulb temperatures. This derating
depends upon the altitude and may amount to as
much as 40 or 50 percent.
To obtain maximum reliability from vacuum
tubes and equipment, it is important that pains be
taken to keep the operating temperature of the
bulb at its hottest spot within the limit specified
by data sheets.
Biography
M. 1. Miller, The Design of Electronic Equipment
Using Subminiature Components, Proc. IRE, Feb. 1950.
B. O. Buckland, Electron Tube Heat Transfer Data,
Elec. Ellg., Nov. 1951.
R. ]. Bibberow, Electron Tube Ratings at Very High
AltitUdes, Tele·Tech., May 1951.
P. T. Weeks, Reliability in Miniature and Subminiature
Tubes, Proc. IRE, May 1951.
D. G. Ko<n, Increasing Tubr: Reliability in Industrial
Circuits, Prod. Eflg., June 1952.
un Subminiawre Tu bes," Sylvania Electric
Products, Emporium, 'Pennsylvania.
"Data on Subminiature Tubes," Raythcon Manufacturing
Company, NewtOn, MasslIChusclts.
R. ]. E. Whittier, Mechanical Considerations Affecting
5 Vacuum Tube Reliability, Part 1, Tele·Te,-h., Feb. 1952.

Ambient °C Bulb temp in °C at hottest spot

23 100 170 230 280 310

160 220 260 300 340 :no
250 310 350 390 420 450

The importance of bulb temperatures on tube

life can be noted in recent information published
by various tube manufacturers showing the life
which may be expected for subminiature tubes.
Most of these tubes are rated for maximum bulb
temperatures of 200 deg. with a few having a
rating of 250 deg. C. A reduction of bulb tem·
perature on the order of 20 percent when operating
in the region of 200 deg. C bulb temperatures will
result in a substantial increase in the life expectancy
of the tubes. The cooling of the tube envelope is
the most important consideration in mounting the
tube.

A loose·fltting shield such as is commonly

employed with miniature tubes may increase the
temperature appreciably. The situation arises be­
cause the shield is not tight fitting but instead
provides a blanket of hot air around the tube. 1 hus
the shield does not provide a good thermal contact
with bulb of the tube or to the chassis and cannot
effectively cool the bulb.

If shields are employed, and they are tight fitting

and can be fastened directly to the chassis, a con­
siderable amount of heat can usually be removed in
this manner. To obtain maximum heat radiation,
the shield should nOt be plated and should not be
polished.

Radiotronics 43 April, 1956

 H e at Transfer in Receiving Tubes
O. H. S c h a d e , J r .

H a rrison

This chapter presents methods for analyzing the heat­ ventional tube deSign.
transfer processes that occur in receiving tubes due to
conduction, convection, and radiation, and for deter­
mining the resultant electrode temperatures. The con­
cepts used in radiant-energy transmission are described
in detail because radiation is the most complicated en­
ergy transfer in a tube structure. A section concerned
with the application of heat theory to tube problems is
included.
HEAT TRANSFER BY STEADY, UNIDIRECTIONAL
CONDUC TION
The Fourier Equation and Two Simple Solutions
Figure 1 . Dimensions of a Solid of Constant
The amount of heat that flows within a body depends Cross -Sectional Area
upon the temperature gradient present, the cross-sec­
tional area of the body, and the thermal conductivity of
the body material. In 1822, Fourier proposed the equa­
tion for simple, steady heat flow by conduction:
q = - KA (dT/ dx) ( 1)
The quantity K is the thermal conductivity of the body,
A is the cross - sectional area, and dT/ dx is the tem­
perature gradient. Actually, the thermal conductivity
of most materials varies with the temperature of the
material, but a mean value Km over the temperature
range encountered usually yields solutions of sufficient
accuracy for most engineering purposes. The solution
of Fourier's equation for a solid of constant cross-sec­
tional area, as shown in Fig. 1, is:
q = KmA ( A T/x) (2 )

Another simple solution involves the flow of heat through

the walls of a cylinder, as shown in Fig. 2, where the
cross -sectional area is proportional to the radius of
the structure. The equation for the heat flow is:
q = Km Am ( A T/x) (3)
A2 -A 1
where Am = .In-;(-'-- - (4)
A2/T7
A 1)
Heat flow is usually expressed in watts, length in cen­
timeters, area in square centimeters, temperature dif­
ference in degrees Centigrade ( or degrees Kelvin) and
the thermal conductivity in watts/ cm2 / deg C for a unit
length. The preceding equation of conductive heat flow
can be applied to most problems encountered in a con-
Figure 2. Dimensions of a Cylinder
The Concept of Thermal Resistance
In many cases, it is convenient to adapt the concept
of thermal resistance in heat-transfer problems . Doing
so permits a network of conducting bodies to be analyzed
in a manner analogous to that used for electrical net­
work analysiS . The thermal resistance may be defined
as the temperature difference divided by the heat flow,
or
R AT/q
= (5 )
When the heat transfer occurs by steady, unidirectional

244

conduction, the thermal resistance may be expressed
(6)
when Eqs . (2) and (3) are applicable. The determination
of the heat flow through a sandwich of several materials
of cross -sectional area A, as shown in Fig. 3, illus­
trates the use of thermal resistances. The total tem­
R3 x 3 / Km3 A. The total heat flow is q (T 1 -T 4)/
perature drop is (T1 -T4 ) . The individual thermal re­
sistances are: R1 = x 1 /Km1 A, R 2 = x 2 / K 2 A, and
=
(R 1 + R2 + R3 ) = L AT/ L R.
=
(2) ( 3)
Figure 3. Heat Flow Through a Sandwich
of Several Materials
The illustration is of a series arrangement of thermal
resistance, but any network of series and parallel ele­
ments can be attacked by the same methods used for
electrical networks. It is also possible to conceive of
"convection resistances" and "radiation resistances"
in the network, as will be illustrated later . The ther­
mal resistances of various materials may be found in
Figs. 15 and 16.
HEAT TRANSFER BY CONVECTION
Convection is the transfer of heat by fluid flow. Most,
if not all, convection problems originate with the trans ­
fer of heat between a surface and a moving fluid. The
physics of the transfer is illustrated in Fig. 4, which
shows a fluid moving along the surface of a solid. Three
types of flow are depicted; a laminar flow at the sur­
face, a buffer layer of laminar flow and turbulent flow,
and a region of turbulent flow at temperature T beyond
the other regions. Heat is transferred through the vis-
Heat Tran sfer i n Receiving Tubes
cous film of laminar flow by conduction and radiation.
Heat is transferred through the buffer layer by con­
duction and convection. In the region of turbulent flow,
heat is transferred primarily by convection.
TURBUL ENT
FLOW
T
Figure 4 . Fluid Flow A long the Surface of a Solid
So many variables are encountered in an attempt to
solve even the simplest problem in convection heat
transfer that boundary layer thickness, fluid viscosity,
density, and heat capacity are assumed to remain con­
stant. If the above-mentioned factors remain constant,
the heat transfer is proportional to the area A and the
temperature drop Ts - T = AT. With the introduction
of a dimensional constant h, Newton' s law of cooling
can be stated:
q = hA AT. (7)
The quantity h is called the "film coefficient of heat
transfer. " It is evaluated by using dimensional analysis
to arrive at an empirical relationship of the pertinent
constants.
Many surface configurations have been evaluated, but
the only problem normally encountered in tube design
is the natural convection from smooth horizontal and
vertical cylinders in air. The expression for h under
0 . 2 7 ( A T/D)O.25 Btu/hr. ft� deg F
these conditions is: 2
h =
or, in more convenient units,
h = 1 .78 ( A T/D) O .2 5 x 10 -4 watts/cm4deg C (8)
where D is the outside diameter of the cylinder.
THE PHENOMENON OF HEAT TRANSFER BY RADI­
ATION THROUGH NON ABSORBING MEDIA
The Physics of Radiation
No theory that explains all the aspects of the phe ­
nomenon called "radiation" has yet been discovered.
245
ELECTRON TUBE D ESIGN
Certain observations are best explained by the concept
of radiation being an electromagnetic wave that is pro­
duced by the change in ene.rgy level of charged molec­
ular and submolecular particles in an emitter. Because
not all particles in the emitter have the same energy
levels, the radiation is composed of an infinite number
of frequencies (and wave lengths) . The distribution of
radiant energy by frequency must be similar to the dis­
tribution of energy levels among the molecules of the
radiating obj ect, and one would expect that the frequency
of radiation varies somewhat like a probability curve
(as will be shown later) .
Radiation occurs at all temperatures and is independ­
ent of temperature difference between radiating bodies .
For example, if two bodies A and B are maintained at
temperatures Ta and Tb, body A will radiate a certain
quantity of heat to body B, independent of the tempera­
ture Tb . Body B will radiate a certain quantity of heat
to Body A independent of the temperature Ta . It is not
possible to "buck-out" the radiation between bodies in
the sense that (in electrical circuits) two batteries of
similar voltage with like terminals attached will not be
able to cause a current. It is possible that the net heat
interchange between body A and body B is zero, but the
force (temperature) that causes energy interchange is
always present (as long as the temperature is above
absolute zero) . This is a statement of the "Prevost
law of exchanges . "
The Black-Body Concept
A body that is a perfect absorber and emitter of ra­
diant energy is called a "black-body. " The concept of
a black body is useful because it provides a standard
to which the absorption or emission of radiant energy
by practical bodies (or surfaces) can be compared.
The characteristics of a black body can be closely
approached by a model first proposed by Kirchoff: If a
small opening is made in a hollow sphere whose inner
surface is covered with a material that is a good ab­
sorber (such as lamp -black), radiation incident on the
small opening will be almost totally absorbed within the
sphere, because of the many reflections and absorptions
of the energy before the much-diminished "wave" again
encounters the opening.
The Distribution of Radiation
The radiation from a black body was predicted by
Wien, based upon the wave theory of transmission. He
proposed that the quantity of energy radiated by a black
body at any particular wave length (called the "mono­
chromatic emissive power") could be expressed
(9)
where C l and C 2 are constants, :\ is the wave length
of the radiation, and T is the temperature of the radia­
ting body . The dimensions of monochromatic emissive
power are radiant energy at a particular wave length,
per unit time, from a surface of unit area. Correla­
tions with experimental evidence were good at short
wave lengths but poor at long wave lengths . However,
246
the wave length of maximum radiation could be deter­
mined with little error from Wien' s "displacement law. "
:\ T = 2880 microns · deg K (10)
max
Planck introduced a corpuscular theory of radiation
which premised that energy is emitted in discrete bun­
dles, or corpuscles, rather than in a continuous form.
He called these bundles of energy "quanta. " The rela­
tionship found between emissive power, wave length,
and temperature, as postulated by the corpuscular the­
ory, agreed well with experimental evidence. The re­
lationship is of the form:
(11)
which is similar to Wien' s law. The solution for the
constants yields the equation:
Eb:\ =3 .73 x 1 0 - 16 :\ -5 /(e 1 . 436/:\.T - 1)watts/cm2 micron
(12)
where T is expressed in de,jrees Kelvin and :\ is in
centimeters ( 1 micron = 10 - centimeters) .
The monochromatic emissive power of a black body
at different temperatures is shown in Fig. 5, as plotted
from Eq. (12) . Notice that as the black body tempera­
ture is increased, the maxima of the curves shift to
shorter wave lengths . The body must be quite hot (about
5000 K , approximately the apparent temperature of the
sun) before the maxima occur near the center of the
visible spectrum.
\
\
\
\
\ 5000 K
\0.577
\
A MAX.=
�
1!I J�����6�0�0�K��==���������
2 4 6
WAVE LENGTH ( A ) - MICRONS
8 �
Figure 5. The Monochromatic Emissive Power
of a Black Body at Various Temperatures
Eb = fa Eb:\ d:\ . The surprisingly simple relation­
The total radiation at each temperature is propor­
tional to00
the area lying under the curve, or by definition,
ship between the total radiation emitted by a black body
and its temperature and surface area was determined
experimentally by Stefan and devised theoretically by
Boltzmann. It is called the "Stefan-Boltzmann law":
(13 )

To obtain the heat transfer in watts, the Stefan Boltz­
mann constant is 6 = 5.67 x 10 - 1 2 watts/cm2 (deg K)4 .
The equation is probably most easily used in the form:
% = 5.67 A(T/ 1000) 4 watts (14)
The total emissive power of a black body is
(15)
The Gray-Body Concept
A body that emits or absorbs with less "efficiency"
than a black body at the same temperature can be de­
fined as a "gray body. " The material (surface) is said
to have an "emissivity" E defined by the ratio
:IT
E = E = constant
At any particular wave length and temperature a body
has the "spectral" emissivity
I
E :\.
E ;I. - E T constant (16)
b :\.=
Furthermore, the gray body has a constant spectral
emissivity independent of temperature or wave length.
The gray body has a radiation distribution exactly the
same as that of the black body (as shown in Fig. 5),
except that everywhere it is some constant fraction of
the black-body radiation. The gray body has a "to­
l
tal" or "thermal" emissivity as defined by the ratio
E 3 . Two continuous black surfaces, one enclosing the
Et - (17)
Eb T constant
=
The gray body has a total emissive power:
E = 6 E t T4 (18)
Fig. 6 is the graphical solution of this equation with E
versus T, and E t as a parameter.
Real materials differ from the ideal gray body in that
their spectral emissivity changes with wave length and
temperature and that their total emissivity changes with
temperature, but for heat transfer calculations, real
materials are usually assumed to be ideal gray bodies.
The Use of Spectral and Total Emissivities
Normally, there is only one instance when the tube
designer need be concerned with spectral emissivity -
when he determines the temperature of a body with a
"disappearing filament" optical pyrometer. Pyrome­
ters of this type are calibrated to read "true" black­
body temperature. The pyrometer reads "brightness"
temperatures on any body that is not black. This oc­
curs because temperature is determined by matching
light intensities; the intensity of the radiating object is
matched to the intensity of the pyrometer filament. If
the radiating object is not a black body, it must be at a
higher temperature than a black body that emits the
same light intensity. For this reason, brightness tem­
perature must be corrected to obtain true temperature;
the correction depends upon the spectral emissivity of
H eat Tran sfer in Receiving Tubes
the radiating object. Since a red filter is normally used
in viewing the body, the spectral emissivity of mater­
ials is normally determined at a "standard" 0 . 66 mi­
crons wave length. Fig. 7 shows the correction re­
quired to obtain the true temperature of a body when
its spectral emissivity is known.
Total emissivity is used whenever the heat transfer
from a body is calculated. When the radiation from a
body is emitted in all directions into or through an im­
aginary hemispherical surface, the term total "hem­
ispherical" emissivity is used. The ratio of total emis­
sive power of the test body to that of the black body,
taken in a direction normal to the test surface, is called
the total "normal" emissivity. Most information avail­
able on materials and material surfaces lists their total
normal emissivity, which can be considered equal to
the total hemispherical emissivity, except for well­
polished surfaces; for well-polished surfaces, the hem ­
ispherical emissivity i s 1 5 t o 20 per cent greater than
the normal emissivity.
SOLUTIONS OF HEAT TRANSFER E QUATION FOR
RADIATION THROUGH NONABSORBING MEDIA *
Black-Body Radiation for Simple Geometries
The heat transfer equations for the black-body ra­
diation of three simple geometries have the same solu­
tion. The geometries are:
1. Small black body radiating to large black enclo­
sure.
2. Infinite parallel black planes .
other; no negative curvature on the inner body.
For each, the net heat interchange by radiation is :
where A 1 is the area of the hotter body.
The View Factor
When geometries become more complex than those
listed above, it is convenient to adopt the idea of a
"view-factor. " Suppose, for example, that there are
two infinite parallel black planes and plane 1 , the hotter
of the two, radiates heat to plane 2 as predicted by Eq.
(19) . If plane 2 is reduced in size, it will receive only
a fraction of the heat radiated by plane 1. The heat in­
terchange will be
where F, the view factor (also called the "interchange
factor, " "geometrical factor, " or "form factor") , rep­
resents the fraction of the heat radiated by plane 1 that
is intercepted by plane 2.
Three "types" of view factor are normally considered.
They are all, by definition, of the form
by body 2 .
F 1, 2 = heat radiated by body 1 intercepted
total heat radiated by body 1
*
E quations and concepts primarily from Rottel , Refs. 4 and 7.
247
ElECTRON T U B E D ESIGN

0:
I&J
�
o
Q.

5 . 6 7 X 10 -12 WATTS
C M 2 DEG
er =

K4
;-;-:::-
TT
-r . ;i; If

TEMPERATUR E - DEG C

Figure 6. Emissive Power as a Function of Temperature for Various Thermal Emissivities

248
 Heat Tran sfer i n Receivi ng Tubes

/ 1 1/ / ./
/ ./ ./
15 // / / /'
lL V v / /

v l/
/ /
I1 / V v
/ / 1/ //
I1 1 A/
10
/
8 / / /
/'
vV
/
6 ���� � /���q-4-4-4-+-+-+-+-+-����������-4-4-+-+�
/ _�
5 L I1 1/
I /. /
I1 l/

flY /
1II1 /
4
3
1

200 400 600 800 200 400 600 800 200 400 600 800
1000 2000 3000 4000
OBSERVED TEMPERATURE - DEGREES K

Figure 7. Corrections to Brightness Temperature Determinations by Optical Pyrometer for Transmission

of Bulb and Emissivity at 0. 665 Microns

249
ELECTRON TUB E D ESIGN
The individual types are:
1 . F: the view factor determined by the geometrical
construction and positioning of two black bodies.
2. F: the view factor,determined by geometrical con­
siderations, which accounts for the effect of other re­
flecting bodies near the two black bodies of concern.
3. -:I: the "over-all" view factor, which accounts
for geometrical considerations, the effects of other
bodies, and the emissivities of two gray bodies of
concern.
The view factors may be evaluated by a mathematical
approach, by graphical integration, or with the help of
mechanical devices. The view factors, F and F, for
several geometrics are presented in Figs. 8 to 1 4 . The
factors apply to:
1 . Adjacent rectangles forming angles of 600, 900,
and 1200•
2. Long, opposed parallel rods.
3 . Parallel plane structures.
4 . Parallel planes and rows of tubes .
The over-all view factor 'J contains the effect of the
emissivities of gray bodies. Only the simplest are ac ­
curately expressed.
Infinite Parallel Gray Planes (F 1). = "right-angled" surface that, for example, might be
= 1
-:1 (li E 1 ) + ( liE 2 ) - 1
A Gray Radiating Surface That is Small Compared to
the Distance to Other Gray Surfaces.
A Gray Enclosed and Gray Enclosing Surface.
F (or F) F (or F)
,...,---,,---
.,--, - ,-,--- <1<:--;--,----;-:-----,-:-...,.....,.
-;-:- ... .,. -,--;--
(liE 1 ) + (l/E 2 ) - 1 (l/E 1 ) + (A ll A2 ) [ ( l/E 2 ) - 1]
The left-hand term is obtained assuming completely
specular reflection of heat (equal angles of incidence
and reflection of a wave) . The right-hand term is ob­
tained by assuming completely diffuse reflection. With
practical surfaces neither extreme is correct, but the
right-hand expression (diffuse reflection) is much more
nearly correct for most surfaces, particularly if they
are rough, oxidized or nonmetallic.
J � ( 1 ) (A A ) (li ) 1
A Gray Body Enclosed by Gray Tubes.
F (or F)
1 El + i 2 [ E 2 - ]
This is a further approximation of Eq. (22).
Pairs of Gray Surfaces Not Forming an Enclosure
(General) .
(2 4)
This covers long opposed parallel gray rods, adj acent
2 50
gray rectangles, gray plane-parallel structures with
boundaries, and the like. The factor in Eq. (2 4) always
tends to give a somewhat low, but reasonably good re­
sult when the surfaces are designed to be good radiators
or are rough, OXidized, or nonmetallic.
In general, the accuracy of calculations of radiation
heat transfer in tubes will be most affected by the dif­
ficulty of predicting the emissivity of the surfaces in­
volved. For example, most metals have a considerably
lower emissivity when their surfaces are clean than
when they are oxidized or covered with cathode evapor­
ants. When the emissivities of these surfaces have
been evaluated and the proper values used in the prob­
lem, the accuracy of the calculations will be truly de­
pendent upon the approximations made in determining
the heat transfer equations.
Radiation from "Composite Surfaces"
It is possible to determine the radiation from surfaces
composed of the more -simple geometrical shapes for
which view factors are known. The radiation from two
of such "comp OSite" surfaces will now be evaluated.
Radiationfrom a "Right-Angled" Surface at Constant
Temperature. Fig. 1 7 shows a cross section of a
found on the half-plates of large beam -power receiving
(20)
tubes. The radiation "loss" from this composite sur­
face to the large enclosure E, which is assumed to be
a black body, is to be found. Body 1 has a radiating
area A 1 (that side facing the enclosure as shown), an
emissivity E t , and a temperature T. Body 2 has a
radiating area A2 ' an emissivity Et (the same as body 1),
and a temperature T; the temperature of the enclosure
(2 1) is TE . Assume, for the moment, that only body 1 is
hot. Assuming diffuse reflection, it radiates the quan­
tity of heat
q = OE t A 1 T 4, of which
(22)
O E A T4 . ( 1 - F ) is lost to the enclosure, and
t 1 1, 2
O EtA 1 T 4 . F 1 ,2 is intercepted by body 2. Of this in­
tercepted heat, body 2 loses
OE t A 1 T4 F 1 2 (l -E d (1 - F2 1 ) to the enclosure and
reflects to body 1 the amount '
(23)
and reflects to body 2
OE tA 1 T4 F 1 , 2 ( 1 - E t) F2 , 1( 1 - E t) F 1, 2
and so forth.
So far, the total losses to the enclosure due to the ra­
diation from body 1 are:
q 1 = OE tA 1 T4 ( 1 - F 1 , 2 ) .:!:. OE tA 1 T4 F 1, 2 (1- E t) (1-F2, 1 ) +
 Heat Tran sfer i n Receiving Tubes

OE tA l T 4 F l " 2 F2 1 (1- E t) 2 ( 1 - F l , 2 ) or Examine the magnitudes of the bracketed quantities. If

the structure is geometrically designed to be a radiator
�
q l =O EtA l T 4 1 - F l 2 ) + F l 2 ( 1 -E t) ( 1 - F2 l ) + F l , 2 F2 , l ' (to have high losses to the enclosure), and has a reason­
( �
l -F l , ) ( l - E t) 2 J ' ably high emissivity, the third term is negligible. For
example, if F l , 2 F2 , 1 0.2 and E t 0 . 5 (normally it
= = =

0.5
\
0.4
1\
Il.

a:
0.3 "-
\t.o PARAMETER r k /r s
1�· 2""
0
�
U

��
<l
Il.
0.2
�:::::-- -
� �
� ......
L.LJ
- ......

=::::::
:>
:---:=:�:::
0. 1 ---=::: 1+----- d ---I

o
o 2 3 4 5
RATIO, dIrk
Figure 8. View Factor for Long, Opposed Parallel Rods

1.0

��
- -f--
-
f--

V t::=t::-
� f--- � :::- - -
....-
j!.
-
0.8 ...--
-- -- - --
-
f--
I--

/ V
I-I- I-- f..- -
v
7 - I-"""
V
- -
-I-
r-
-��

/ ./V ././" J--

...-- ...-- --
./" VV V I-I-
I- �f--
-
6 -f--
V
./' .-/ -I--
-' ...-- --I--

V 1/ VV )-
,-
I--"""" !----
V V
..- - -
V
...-- - --
- -

V""
/"'
./" V l-

V
./'
V V
V
!5
Ill.
0.6 -/

/ / /V /
V

>
./
/ / /'/ ./ ..-V
./'

V V V /'
V
VVV
/ J '/
a:
If

7V
o
�

�
()

�V(
/ V

/, i/ / /
0.4 // 'i V /
I/ / /
�
L.LJ
RADIATION BETWEEN PARALLEL

I fiJ V / /t/
:>

�V
PLANES, DIRECTLY OPPOSED.
1-2-3-4 DIRECT RADIATION
0.2 VI / I �;/
BETWEEN PLANES, F

�V
5-6-7-8 PLANES CON NECTED BY

I�V
NONCONDUCTING BUT
/ / RE-RADIATING WALLS, F

r;,V �V
I 1-5 DISCS, 3-7 2:1 RECTANGLE
V 2-6 SQUARES, 4-8 LONG, N ARROW
RECTANGLES

o
�� 2 3 4 5 6 7
RATIO - SMALLER SIDE OR DIAMETER/DISTANCE BETWEEN PLANES
Figure 9. View Factor for Opposed Parallel Disks, Squares, and Rectangles

251
ELECTRON TU B E D ESIGN

is even higher than this) , the bracketed terms have the
values
[
0.8 + 0.08 + 0.008
and the third term is surely negligible. Therefore, let
the radiation loss from body 1 to the black enclosure of
F 1 3 is determined by the heat interchange in plane­
parallel structures, while the view factor F 1 2 is deter­
] mined by the heat interchange between adj acent, per­
pendicular planes.
Assuming diffuse reflection with only body 1 hot, body

O E tA 1 T 4 �� 1 - F 1 , 2 )+F 1 , 2 ( 1 - F2 , 1 ) ( 1 -E t ) ,
the quantity be defined as 1 radiates

q1 = ] O E t A 1 T4 ( 1 - F 1 , 2 -F 1 , 3 ) to the enclosure,
6E A T4 (F ) to body 2 and
when only body 1 is hot. Now, if only body 2 were hot, t 1 1, 2
it would have a loss of the same form, or

[
q2 = OE tA2 T 4 ( 1 - F2, 1 )+F2, 1 ( l - F 1, 2 ) (1 - E t) , ] The radiation distribution at body 2, due to the heat
from body 1, is
which can be obtained by a similar method of attack.
The total loss, with both body 1 and body 2 hot, can be
factored to the form
{ [
q = q 1 + q2 = O Et T 4 (A 1 + A2 ) 1 + F l, 2 F 2, 1 ( E t - 1 �
OEt A 1 T 4 F 1 2 ( 1 -E t) ( 1 - F2 " 1 -F2 3 )
- E t (A 1 F 1, 2 + A2 F2, 1 ) } closure. '
lost to the en-

When the large black enclosure is also hot, the net heat The radiation distribution at body 3 due to the heat from
transfer to the enclosure is body 1 is
6Et A 1 T 4 F 1 , 3 ( l -E t) F 3 1 reflected to body 1,
,

If the radiating surfaces are black, E t = 1 and

Further tracing of the he�t reflections leads to loss

components bearing (1 - Et) terms, which may be neg­
indicating, as one would expect, that there are no re­ lected if the same assumptions are made as in the pre­
flections between the radiating surfaces. This expres­ ceding case. The total radiation loss to the enclosure
sion may be further simplified by recognizing that there due to body 1 is then
can be no net heat interchange between black bodies at
the same temperature, independent of their geometries,
so that

This equation further simplifies the black "composite" Since body 3 is geometrically similar to body 1, it has
surface radiation to the same heat loss to the enclosure. The heat loss from
body 2 is evaluated in similar fashion. The total loss
to the enclosure is

where it is necessary to determine only one view factor.

When Eqs . (2 5) , ( 2 6) or (27) are applied to a practical
structure such as a plate, an average plate temperature By geometry, certain view factors have the same value:
and enclosure temperature should be used. The view
factor may be obtained from Fig. 12
A U -Shaped Radiating Surface at Constant Tempera­
ture. Fig. 18 shows a U-shaped surface radiating to a
large black enclosure . To find the radiation loss from
within the U, a method may be used similar to that used
for the right -angled surface. The total loss from each
body, including reflection losses, is calculated assum­
ing that only one body at a time is hot. It is necessary By summing, simplifying, factoring, and accounting
to consult two view factor figures to use any equations for the temperature of the large black enclosure, it is
that might be derived. For example, the view factor found that the net heat interchange is

2 52
 Heat Tra nsfe r in Receivi ng Tu bes

{ [A
q = 2 6Et (T4 _T E 4) ( 1 - F 1 , 2 - F 1 , 3 ) 1 + ( 1 -E t)(A 1 F 1 , 3 equality of heat interchange between black bodies at the
[ J}
same temperatures requires that
+A2 F 2 , 1 ) + ( 1 - 2 F2 , 1 l (A2 / 2 ) + ( l - E t) A 1 F 1, 2 ( 2 8)

Ifthe radiating surface is black, E t = 1 and Eq. (28)

reduces to This equality may be applied to E q. (29) to further sim­
% = [
O (T4 _ TE 4 ) 2A l ( 1 - F 1 , 2 - F 1 , 3 ) + A2 ( 1 - 2 F 2 , 1 ' � plify it, if desired.
THE APPLICATION OF THEORY TO HEAT TRANSFE R
(29)
PROBLEMS IN TUBE STRUCTURES
indicating that there are two surfaces of area A 1 losing
the fractional heat ( 1 - F 1 2 - F 1 3 ) to the enclosure, Heat Transfer Circuits of the Diode
and a surface A2 losing the frac'tional heat ( 1 - 2 F2 1 )
to the enclosure. As for the right-angled surface, the It is often convenient to think of electrical circuits,

X = b/c
Y= a/c

1.0
I--I-- 10.0
��
4.v

.,,-I:?':V
l.---::: V
0.7
-- I-- 2.0
P
I-- ,..-
� V" V V
0.5

� ��
�
1.0
V-

� �
I-- f-
V V�

� � /'
0.3
/. V I ..... � 0.6

VV V
V -�
I--
1/
�
0.2 0.4

V
/ vV' / I--+--

�V
-

/ /
/ V V Vi-""

�u
V
V /'
a::
� V
/. / /
O. I
� // / / V -
f..-f-
V/ /
V V V /
/

W'/ 7
..
....""--

//// /
0.07 /'
/ / �
�

/ /
w
:>

v/ V
V
0.05 O.I = X
V/ /
/
V I--,..-
V

:/ V / /
V
/ // �
----

/ /
0.03
/

V V
V V
/
V
0.0 2
// /
V
/

/
V /
V
0.0I
0.1 0.2 0.3 0.5 0.7 1.0 2 3 5 7 10 20
Y

Figure 10. View Factor for Directly Opposed Parallel Planes

2 53
ELECTRON TUBE D ESIGN

or networks, that would produce current flows similar
to the heat flows encountered in tubes. It then becomes
a simple matter to understand the types of heat flow
which might be encountered in a tube problem and to
appreciate their effects on electrode temperatures re­
sulting from the magnitudes and directions of the heat
flows. Fig. 19 shows a heat-transfer circuit of a di­
ode. The heater is at "potential" TH with respect to
the ambient TA , and forces heat to flow through the var­
ious resistances shown. The temperatures of the cath­
ode, plate, and enclosure are designated Tk, Tp , and
TE , respectively. It is assumed that the enclosure is
at a constant temperature for the purposes of this ex­
ample, and that there are no temperature gradients on
the various electrodes. Heat flow through any insula­
ting spacers (micas) has been neglected. The resist­
ance r o is not of concern because the heater tempera­
heater coating (normally a very small amount) and q
the radiation transfer. At the cathode, heat is trans­
ferred by conduction through a cathode tab, r 5 (norm­
ally very small), and radiation to the plate through r 6 '
The plate loses heat to the enclosure by stem-lead con­
duction through r 7 (which in " high-dissipation" struc­
tures can usually be neglected) and radiation through
r 8 ' The enclosure loses heat to the surrounding me­
dium by conduction, convection, and radiation through
r 9 , r l O , and r ll ' The quantities normally of interest
are the temperatures TH , Tk , Tp , and T E . Tempera­
ture Tk can be measured accurately with an optical
pyrometer; T E can be measured with a thermocouple
or "Tempilac, " or it can be estimated by using Fig. 34.
T H is quickly approximated (roughly) by using the
Kauzmann nomograph for heater design or, more ac ­
curately, by referring to date on the resistivity of tung­

circuit of iJlg. 20 may be used. The normal procedure

ture and power input are usually known or easily cal­
culated. The total heater power divides among the
resistances r 1 , r2 , r 3 , and r 4 ' The resistance r 1 is
the thermal "conduction" resistance of the heater legs,
which in most tubes is high enough so that any heat
flow through it is relatively small. The heat flow
through resistance r2 is the radiation "end loss" from
the heater at the ends of the cathode. Heat is trans ­
ferred between heater and cathode through resistances
r3 and r 4 , r 3 representing the conduction through the
sten at different temperatures. To determine the tem­
perature T , with Tk and T E known, the simplified
is to assume an average plate temperature and to cal­
culate the net heat transfer between cathode and plate
and between the plate and enclosure. These two heat
transfers will be equal when the proper assumption for
Tp has been made. !f an electrical dissipation appears
at the plate, the two heat transfers will not, of course,
be equal. In this case, the plate will assume a tem-

N = o /b
L = c/b

V
O. I = L

/'
....-
V I-l-
.--- 0.2

/'"
v�
/ 0. 4
--

/
V ,/V
/
;;;;;-
/ -
0.6
/ / ....V
/ V
..

/ /
1 .0
/ / ,/
--
--- ----

/ /
/ V
V

/ ....1/'
..
/ ....V
.. -f--
2.0
/

/ V
I--"
/ ....- �

V V
V
/' './V 4.0

V ./V V ��
V V V I--
V V 6.0

I---" l.---j.....--
...... -
I--
........- V -
10.0
I--" f--
I--� I-I-�� �I-- I---
"""
- - -
I- l---- +--;-
o
0. 1 0.2 0. 3 0.5 0.7 1 .0 2 3 5 7 10 20
N

Figure 1 1 . View Factor for Direct Radiation Between Adjacent Rectangles (fJ = 60°)

2 54
 Heat Tra nsfer i n Receiving Tubes

perature such that it loses the amount that it intercepts or, if the enclosure is glass, where the assumption that
from the cathode, plus the electrical dissipation. The E E = 1 may be made,
equations for radiation heat transfer between cathode
and plate and plate and enclosure are of the form q = 0A- -p
(Tp 4 - TE4 ) p F
E (32)
If the plate surface is geometrically reasonably
smooth, that is, if it has only small portions of its sur­
where '7 is normally described by Eq. (22) . The equa­ face capable of "viewing" other portions, F in Eq. (32)
tion for cathode-to-plate radiation is: is unity.

[
F
q - 0 Ak (Tk 4 - Tp 4) (30) To avoid complications, it is usually convenient to
( k ) + (Ak/Ap) ( l/ E p ) _ �
l / E use the coated area of the cathode in radiation calcula­
tions . This assumption does not normally produce ser ­
where F = 1 if the cathode is completely surrounded by ious errors in calculations requiring the usual engin­
the plate (the usual approximation) ; for plate to enclo­ eering accuracy. If the plate is assumed to be con­
sure radiation, the equation is : siderably hotter than the enclosure, such that
F
q - OAp (Tp4 - T E 4)
( l/E p ) + (Ap/A E ) [( 1/E E ) - l
(3 1)
] Tp4 » TE 4 ,

N = o/b
L = c/b

0. 5

..",.I.--
O. I = L

I- j.---
I-
�I--

V
0.4 0.2
V
� I-
.. V
....V

l/ ..... ...- ---

V
./
v.4

/ V -

/ ....�
.-
LL..

/ /'
0.3 0.6
V V

.--�
/'
0:: --
0
V
I-

�
U
V
1.0
V /'V
«
-

/
V
LL..
�
V ./
/
:>
ILl

/ V
0.2
V

.--
./

/
2 .0
/' ..
....V

..... ....�
.-
�-

IV / ....V ...-
/

V
..
V VV
4.0

V...... ...-� -�
Vv ��
O.

..- j.---
�� �I--
6.0
f--
--
- JO.O
v � I-- i-- -
-
-

o
0.1 0.2 0.3 0.5 0.7 1.0 2 3 5 7 10 20
N

Figure 12. View Factor for Direct Radiation Between Adjacent Rectangles (jJ = 900)

255
ELECTRON TUBE D ESIGN

then the radiation from a relatively "smooth" plate can

be quickly approximated by reference to Fig. 6. The
l'
transfer through r The total heat the plate must lose
through r 5 and r 6 to T E is then composed of the radia­
heat conducted from the plate by a stem lead will be tion through ri, and the plate dissipation. The average
discussed later. plate temperature may then be solved for by the same
method used in the diode analysis. The radiation from
Heat Transfer Circuits of the Triode cathode to plate differs, however, from that of the diode
because the grid intercepts a fraction of it, Fk l (as may
The heat transfer circuit of the triode is shown in be determined from Fig. 14) . The expression for the
Fig. 2 1 . The temperatures Tk, T s ' and T E are as­ cathode -to-plate radiation for the triode is
sumed to be known, so that the Simplified circuit may
[ ( l IE
be used. Heat is radiated from the cathode to the plate F· (1 - Fk1)
q = 0 Ak (Tk 4 - Tp4 ) l / (33)
through r 1 and from the cathode to the grid through r 2 '
The grid loses heat by conduction through r 3 to the stem
( Ek ) (Ak/ Ap)
+
p ) - lJ
at temperature T s . It may either lose or gain heat where the usual approximation, again, is that F 1.
through q, depending upon whether Tp is higher or =

lower than TC' As in the case of the dlOde, the plate

loses heat to the enclosure by radiation and conduction Once the average plate temperature Tp is established,
through r 5 and r 6 and, similar to the grid, may either the average grid temperature can be calculated. The
lose or gain heat by radiation through r 4 ' To avoid an simplified heat transfer circuit is shown in Fig. 2 2.
over-complicated trial-and-error solution for the aver­ The procedure again involves a trial-and-error solu­
age plate and grid temperatures, an approximate circuit tion, assuming different values of Tg until the radiated
analysis will be presented. It will be assumed that all heat through r2 is equal to the algebraic sum of the
the radiation incident on the plate is due to the heat conducted and radiated heat through r3 and r 4 ' The

N = a/b
l = c/b

0. 1 = L
--
I-""
0.20
�
V
0.2
-
-f-"
V
./

/'
//
/
0.4
� 0.1 5 ./
/V" �!--
/

/'
0.6
a:: / ......
,/
......
o
I­ /" ..---

/' /
u ,/
Lt
......-
./
1 .0
/ /" --
;t 0. 1 0 -
// --
/' ....�
w

V
./ ..-
/' V
> 2 .0
./" ./" / -
--
0.0 5 � / /" -- 4.0
--�

�-
� ......
.....-

� V -- -
6.0
- -� - -I-"" --- 1 0.0
I---- -� - - -
o
0. 1 0.2 0.3 0. 5 0.7 1.0 2 3 5 7 10 20
N

Figure 13. View Factor for Direct Radiation Between Adjacent Rectangles (fJ = 1200)

256
 Heat Tra nsfer i n Receiving Tubes

cathode-to-grid radiation through r2 can be expressed whe re Fp1 = F 1p = Fk1 . At this point, it is necessary
Fk1 that the heat conducted from the grid be evaluated to
4 4 complete the analysis.
q = O Ak (Tk -Tg )
(34)
( l / E k) + (Ak/Ag) [ ( l / E g) - lJ
Heat Transfer by Conduction Through the Grid and

"finishe8 - length" times the "length of a turn of wire ".

where A is the area of a solid grid plane of dimensions
If the grid is hotter than the plate, the grid - to-plate
radiation will be
where F 1p = Fk1 . If the plate is hotter than the grid,
the radiation interchange is
Stem Leads
It is possible to determine the temperature gradient
or the grid, and an average grid temperature due to
conduction heat flow by considering the grid to be a net­
work of thermal resistances. Two cases will be treated:
one in which a stem lead is welded to each side rod and
heat flow is symmetrical about the grid center line, and
the other in which there is a stem lead welded to only
one side rod.
Stem Lead to Each Siderod. Fig. 2 3 shows the ther­
mal resistances in one side of a grid structure which
has been divided into five groups of lateral wires. Each
group of lateral wires is composed of three resistances
r on one side of the center line, about which fractional

0.8
1\ K = bId

\
\ � � 1 "t".(
�
0.7

�
ROW 2

� � '1
1\
ROW I

�
0.6
\
�

RADIATING PLANE
0.5 1\
\
""
a::
o
t-
� 0.4
""
IJ...

:;
� ROW I
W

""
'"
0 .3

I-- --
V-t--r--� 2 .......
0.2 J

/
r-- -
r-- --
r---r-- r---t--
r---
-

I
t--- -

O. I
/
11
2 3 4 5 6 7 8 9 10 11
RATIO K

Figure 14. View Factor for Rows of L ong Tub es

257
N m
U1 r-
(X) m
()
-t
;;:c

�--B::-l- l-±:E::i::HllittW:-:R--: --:-iT...+£l-T1':I �tn±!I 100

_'_I-.'"-.'-Ft=:::EE__fit:j:t:: ijt!:ltt�8=5=t!:E:Et5o��
0
Z
-t
60 60 C
0'
m
::::---t.:=t!:t:lli:t��.pt�:: =H��
�'tl-=-� �t�T
m 0
30 30 m
�
G)
___ .'.''''='C �.c�'�.:...u.l. jJIU Z

10 10
I
(J)
...J

i 6 6

�
-
=:-g :=-,-
A NICKEL ON OFHC COPPER­
I
a:: * 27%
=-
w 3 -- - -
I- - -t - BY
CROSS -SECTIONAL AREA 3
w
::E - . - * 40% OFHC COPPER ON COLD -
1
,
S 2
·- - "·- " '.
c STEEL - BY WEIGHT

1.0

0.6

0. 3 Ejc�= 0.3

0. 1 0.1
10 3 6 100 3 6 1000 3 6 10,000 3 6 100,000 3 6 1,000,000
THERMAL RESISTANCE PER CM OF LENGTH - DEG C/WATT/CM

Figure 15. Thermal Resistance of Round Wires (300 C)

 H ea t Tra nsfe r i n Receivi ng Tubes

1000

(f)
...J 6
:::!:
w
0::
ex

I
::>
0
(f)
3

ex
w
0::
ex

H-f+

6 �--H- +
--, • l '
--l

,-'
:.:: '* C A LC U L AT E D F R O M 2 7 % A- N I C K E L ON OFHC COPPER -

�
3 ===== BY CROSS SECTIONAL AREA .

� NOTE D .
£
CONDUCTIVITIES TA K EN AT 300 C , UNLESS OTHERWISE
r-

1 0 �U.��ww��������������������WllWID���LU�W
10 3 6 100 3 6 1000 3 6 1 0,000
THERMAL RESISTANCE PER CM OF L ENGTH
DEG C/WAT T/CM
Figure 1 6. Therma l Resistance of Some Tube Materials

heat inputs q are assumed to be evenly distributed over so that

the grid surface. These fractional heat inputs must
flow through the lateral - wire-group resistances to the
siderods, where they j oin the heat from the wire groups where R L is the thermal resistance of a stem lead.
on the other side of the siderods and flow down through There is an additional temperature rise from the stem­
the siderod and stem lead to the glass part of the stem lead-to-siderod junction at T L to the bottom of the first
at temperature Ts ' If T s is assumed to be known, the lateral wire group of the quantity
temperature rises from the glass to the wire groups , 3 0 q Rx '
as well as the temperature rise along the wire groups, where Rx is the siderod resistance from the junction
can be calculated . to the bottom of the lateral wire. The additional rise
to the average siderod temperature Tsr is
The temperature rise along the stem lead is Rf Rf Rf
L 3 0q -.1 . 24q+ � . 18q ,
(2. 15 q) R L , +
10 5 5
259
ELECTRON TUBE D ESIGN

where Rr is the resistance of the "finished length" of The preceding material provides the solutions for the
L heat conducted from a grid. These solutions are nec­
the grid siderod. The temperature Tsr may then be
expressed, essary for the average grid temperature calculations
Rf Rf Rf mentioned in the paragraph on heat transfer circuits of
L L L the triode .
T sr T L + 30qR x + 30q -- + 24q -- + 1 8q .
10 5
= --

5
The temperature rise through the third lateral wire
group to the center line of the grid is
3qr + 2qr + qr = 6qr.
If the average grid temperature is considered to be the BODY I ---�
average temperature of this group of wires,

By definition, the total heat to be conducted from the

grid is
qc = 60q

The geometry of the network chosen yields the relation­ Figure 1 7. Cross Sectian of a "Right-Angled" Surface
ship between the fractional lateral wire resistances r
and the resistance of one finished turn of wire R t ,
Heat-Transfer Considerations for Multigrid Tubes
r = ( 1 / 12 ) Ht · (5/N)
The complete solutions for heat transfer problems in
where N is the number of finished turns on the grid. multigrid tubes would be extremely tedious without using
These quantities may be substituted into the expres­ some simplifying assumptions. As in the case of the
sions for the average siderod temperature and the aver­ triode, the cathode-to-plate radiation can be assumed
age grid temperature, yielding to be the only incident radiation on the plate. The equa­
tion for this transfer assumes the form
[
Tsr = T L + qC (Rx/ 2 ) + ( 1 9/ 1 00) R fL J (3 7) ( 1 - Fk 1 - Fk2 - Fk3 . . . - Fkn)
and
Tg = � 8
T L + qC Rx/ 2 ) + ( 1 9/ 1 00) Rr L + (Ht/48N (38)
q CAk(Tk4 _ Tp 4)
=

[
( 1 / E k) + (Ak/ �) ( 1/ E p) - l J
(4 1 )

Usually, the accuracy of calculations is not seriously

Notice that qc is the total heat to be conducted from the
grid, which is not necessarily the total heat intercepted
by the grid. The temperature T L can be obtained by
measuring (or estimating) Ts and adding the tempera­
ture rise along the stem lead, or by referring to stem
heat-flow characteristics such as shown in Figs . 2 4 to
32. These characteristics indicate the temperature of
a "reference point" on the stem lead a certain distance
from the fillet and at the inner fillet, as a function of
the heat flowing through the lead. In this case, the
quantity qc/ 2 flows through one stem lead. Linear in­
terpolations to find temperatures at intermediate points
on the lead are permissible .
impaired by neglecting radiation interchange between
grids. In beam tetrodes and pentodes, the effects of
beam-confining electrodes and suppressor grids can
usually be neglected when calculating heat interchanges
between control grid and screen grid and the plate,
since their view factors are normally small. It is im­
portant to remember that any electrical dissipation on
a grid must be added to the intercepted radiation to ob­
tain the total heat that must leave the grid. If one grid
radiates through another to the plate, the equation is of
the form

Stem Lead to Only One Siderod. A network of resist­

ances similar to that depicted for two siderods may be
constructed. Assuming the "floating" side rod to be
thermally absent from the structure, all the heat to be
conducted must pass through the entire length of the
lateral wires and down one siderod. The resultant tem­
since the intermediate grid intercepts radiation (which
can be neglected in determining its temperature), and
where F 1 p = F k 1 and F 2p = Fk2 . If the plate radiates

T sr = T L + qc [ R x + (38/ 1 00) Rr L J
peratures are : to one gria through anotlier, the equation is of the form

( l/E p) + (Ap/ Ag 1 ) [ ( 1/E g1 ) - l

F 1 ( 1 - F 2)

Tg = T L + qC [ R x + (38/ 1 00) Rr L + (Ht/ 1 6NU

(39)
q = oAp (Tp 4 _Tg l4 ) P P (43)
(40)
J
In this case, all the heat qc flows through a stem lead.

260

While it may seem that the accuracy of the calculations
is questionable in the case of multigrid tubes, the auth­
or 's experience has been that the assumptions made do
not seriously affect the results. For example, the grid
and plate temperatures in the 6L6G were calculated for
the operating conditions used in a test made by Power
and Wittenberg. 8 The actual temperatures were meas ­
ured by thermocouples. The results were:
Measured Calculated
415 C 425 C
292 C 300 -330 C
515 C 525 C
The temperature of the control-grid siderod was ap­
proximated because of the effect of a grid radiator .
It was assumed that the radiator operated at the aver­
age siderod temperature. The accuracy of the tem­
perature calculations that the author has made for con­
ventional tube structures, using the proposed solutions,
was usually better than ::: 10 per cent.
EXAMPLE OF HEAT TRANSFER CALCULATIONS -
THE 6L6G P LATE AND SCREEN-GRID TEMPERA­
TURES
The following examples illustrate possible solutions
of the heat transfer equations for determini ng type 6L6G
plate and screen-grid temperatures.
Operating Conditions: Ef = 6.3 volts, Pf = 5 . 67 watts
�2 = 1 . 25 watts
BODY I
BODY 2 --� � peated. The final value for Tp will be close to 425 C .
RADIATION " LOSS"
Figure 18. Cross Section of a " u" Surjace
Solution for the Average Plate Temperature
The simplified heat transfer circuit for the deter­
mination of the average plate temperature is shown in
Fig. 3 3 . Because the plate temperature lies between
that of the cathode and that of the enclosure, heat flow
q 1 plus the 18 watts of plate dissipation must (by a nodal
Heat Tran sfe r in Receiving Tubes
analysis of the circuit) equal the sum of heat flows q2
and q3 . Because the heat flow through the stem lead
q3 is in the order of tenths of a watt (which is small
compared to the 18 -plus watts that the plate must lose) ,
it will be neglected . The problem will be solved by as ­
suming different plate temperatures and calculating the
heat interchanges until the condition
q 1 + 18 = q2 (watts)
is met.
The heat interchange equations are:
(1 - Fk 1 - Fk2 )
q 1 = a Ak (Tk 4 TP 4)_ (from 41)
(l/E k) + (Ak/ Ap) [(l/E p) - l]
and
(from 3 2 )
The values of the different parameters are:
Tk 1060 K (true temperature)
=
Te 1 60 C = 430 K (from Fig. 3 5, which gives a
=
hot- spot temperature of 185 C for a ST- 1 6 bulb with
pressed stem at 25 watts input; a slightly lower aver­
age temperature is assumed)
E k = 0.30
EP = 0.85
Fk 1 = 0. 1 6 (from Fig. 14; 32 TPI of 3 . 3 - mil wire)
Fk2 = 0 . 13 (from Fig. 1 4; 32 TPI of 3 . 6- mil wire)
Ak = 1 . 7 cm2
Ap = 2 1 cm2 (assuming a smooth surface)
If theaverage plate temperature is assumed to be 400 C
(670 K), then q 1 = 2 . 17 watts, and q2 = 2 . 17 + 18 = 20.2
watts . If this value is substituted into the equation for
q2 , the resulting value of 69 5 K (42 5 C) is obtained.
This value agrees reasonably well with the assumed
400 C. If a closer agreement is desired, a higher value
than 400 C should be chosen and the calculations re­
Solution for the Average Screen- Grid Temperature
A simplified heat transfer circuit for determining
average screen grid temperature is shown in Fig. 34.
It will be assumed that the grid is hotter than the plate
so that, by nodal analysis,
q l + 1 .25 = q2 + q3 (watts)
The heat-interchange equations are:
( /E k) + (Ak/Ag2 ) [ ( 1/E g2 ) - 1]
Fk2
q 1 = aAk (Tk4 - Tg2 4) (from 34)
l
(1/E g2 ) + (Ag2 /Ap) [ ( 1 /E p) - l]
F2 P (from 35)
q2 - a Ag2 (Tg2 4 - TP 4 ) .,.--,_...,.....,.
- .- _ .- ,---.,...,.,.
- _ .., ---,---:::-
261
E l E CTRON TUB E D ESIGN

and

where
1
The values of the parameters are: Figure 20. Simplified Heat Transfer
Circuit of a Diode
Ak = 1 . 7 cm2
Ag2 = 8 . 6 cm2
Ap 2 1 cm2
=

Tg Tp
Tk = 1060 K
T P 69 5 K (from the preceding paragraph)
=

Ts = 150 C (estimated)
Fk2 = 0. 13
F2p = 0. 16
Ek = 0.30
E P = 0.85
E g2 0 . 85 (carbonized)
=

R L = 1 1 00 C/watt ( 1 1 mm of 20 -mil nipron)

R x = 2 1 6 C/watt (9 mm of 40-mil nipron)
Rf 4 = 875 C /watt (3 6.5 mm of 40-mil nipron)
R t = 23, 500 C/watt (23 . 5 mm of 3 . 6-mil moly wire)
N = 46 Figure 21 . Heat Transfer Circuit of a Triode

Assume Tg2 = 800 K; then q l = 0.3 16 watts . The sum

of the screen-grid dissipation and ql is, then, 0.32 +

1.25 = 1 . 57 watts . The value of the grid-to-plate radia­

tion of q2 = 1 . 1 1 watts , and q3 = 0.45 watts; their sum
is 1. 56 watts, which is close enough to the value ob­ Tp

1 J
tained for q 1 to consider the 800 K (530 C) assumption
valid.
I
Figure 22. Simplified Heat Transfer Circuit
of a Triode
LIST OF SYMBOLS
q heat flow, heat interchange
% black -body radiation
qc heat conduction in grid
E total emissive power
Eb total black-body emissive power
EA monochromatic emissive power
Eb A monochromatic black-body emissive power
TA
�------���� A wave length
T temperature
Figure 19. Heat Transfer Circuit of a Diode Te average enclosure temperature
If desired, the average siderod temperature may be Tg average grid temperature
calculated from the expression: Th average heater temperature
Tk average cathode temperature
TL siderod-to-stem- lead junction temperature
The result for the 6L6G is 525 C. Tp average plate temperature
2 62
 H eat Tra nsfe r in Receivi ng Tubes

�q �q 300 .-----------------------.---.----,__�
�q
0.030- INCH NICKEL- PLATED
COPPERWELD INNER LEAD

2X3q� q �q q
� � (J
Cl 200 r---�----�--�
w
o
2X6 q� �q �q
TSR -- W
0:
:::l
f-
2X9q� �--�,�A,r-
�q ----� q �q � 100 I-----t---
� w
a..
:!:
� 50 f---,;''+--::� ' �--,
2X12q �q
�
o 0.1 0.2 0.4 0.6 0.8
HEAT FLOW THROUGH LEAD -- WATTS

I
I

CENT ER LlNE-----j
Figure 25. Stern Heat-Flow Characteristics jar

I
I
1 -1/4 -Inch Button with Niclzel Plated
I Copperweld Inner Leads

300
Figure 23. Thermal Resistances in One Side 0.030 -INCH KULGRID INNER L EAD
oj a Grid Structure (CALCULAT ED)
(J
300 r-------�--�--r_�� Cl
0.030 - I NCH N ICKEL
INNER LEAD � 200 �--�----�--�
I
w
u
0:
:::J
Cl
w f­
Q 200 �----4----+--+' e:(
w
I ffi 100 1-----_+_- .Q/+-----j---_±___.
a..
a::
:J
:!:
f­
w
e:( f-
a::
� IOO, r-----���_+--�--�� p
::!:
w o 0.2 0.6 0.8 1 .0 1.2 1.4 1.6
f- 5011---.....,."'---t----�-"� HEAT FLOW THROUGH LEAD-WATTS
Figure 26. Stern Heat-Flow Charac teristicsjor 1 -1/4 -
o 0.05 0.1 0.2 0.3 Inch Button with 0. 030-Inch Inner Leads (ca lculated
HEAT FLOW THROUGH LEAD-WAT TS jrom the characteristic jar 0. 040-Inch Kulgrid inner­
Figure 24 . Stem Heat-Flow Characteristics jar lead stern)
1 -1/4 -Inch Button with Nicke l 1111le r Leads
300 .------,---,---,
0.030-INCH N I CKEL INNER LEAD
WITH 0.040-INCH KULGRID CONNECTION
(J
average stem (glass ) temperatur e

I
Cl
average grid- side rod temperature W
o 200 I------+--f--_+_-
spectral emissivity
total emissivity W
total emissivity of enc losure 0:
:J
�
total emissivity of grid ffi 100
total emissivity of c athode a..
:!:
w
total emissivity of p late f-

thermal conducti vity

o 0.2 0.4 0.6 0.8 1.6
mean thermal conductivity HEAT FLOW THROUGH
thermal resistance
Figure 27. Stern Heat-Flow Characteristicsjor 1 -1/4 -
thermal resistance of the "fi nis hed" grid Inch Button with 0.030-Inch Nickel Inner Leads, 0. 040-
length Inch Kulgrid Leads Welded to the Nickel at a Point
thermal resistance of a stem lead 5 Millirneters Above the Fillet

2 63
ELECTRON TUBE D ESIGN

300 ,-----,---�---�-- Am mean cross-sectional area

0.040- I NCH KULGRID Ae enclosure surface area
INNER LEAD
u Ag "solid-surface" grid area

I
c.!)
w Ak cathode -surface area
Q 200 �----�-----_+-
Ap plate-surface area
w F, F, '1
a: view factors
;:j
�
x length
5 100 D diameter
a..
:::!i Cl Stefan-Boltzmann constant
� 50 ��
��r_----r_--��--�- film coefficient of heat transfer
constants.
o 0.5 1.0 1.5 2.0 2.5
HEAT FLOW THROUGH LEAD -WATTS
300 r------.--.-�,___,
Figure 28. Stem Heat-Flow Characteristics jor 0.025- INCH KULGRID INNER LEAD
1 - 1/4-Inch Button with 0. 040-Inch Kulgrid Inner Leads 9 - PIN MIN IATURE
u

1 200
c.!)
w
�--+---+---�-
300 ,-------,---�--�
FSI419F PRESSED STEM TEMPER­ w
a:
u
ATURE
�-----,----+---:::� AT 6 M M. ;:j
POINT �
1
c.!)
w
200 r-----r----4�� 5 100
a..

� 50 h�4--__1f___--+-_1
:::!i
w
a:

�
;:j

a:
� 1
00 �----1-�7L-+------r----_��������+T-E-M-P-E-R�
- o 0.1 02 04 0.6 0.8
�
�
ATURE HEAT F LOW THROUGH LEAD -- WATTS
AT PRESS Figure 31 . Stem Heat-Flow Characteristics JOY 9-Pin
Miniature Button with 0. 025 -Inch Kulgrid Inner Lead
o 0.1 0.2 0.3 0.4 0.5 0.6
FLOW THROUGH LEAD AND SPUD --WATTS RE FERENCES
Figure 29. FS1419F Pressed Stem Heat-Flow HEAT TRANSFER BY STEADY UNIDIRECTIONAL
Characteristics CONDUCTION
1 . Giedt, W. E . , Principles of Engineering Heat
300
0.020- INCH NICKEL INNER LEAD
9-PIN MINIATURE
I. �V Transfer, D. Van Nostrand, 1957
2. McAdams, W. H. , Heat Transmission, 3rd Ed. ,
u
"
�o Mc Graw-Hill, 1954
tB
1 200 ���. 3 . Obert, E . F . , Elements of Thermodynamics and
Heat Transfer, 1st Ed. , McGraw-Hill, 1949
w
a:
;:j
��
l..�<'y� HEAT TRANSFER BY CONVEC TION
�
� �
V
a:
� 100 See Refs. 1, 2 , 3
� IQI'" 'AI 1oIl n l �
,,�
V
••0 :RAI
I'" r..n 1:
:::!i
w
50 THE PHENOMENON OF HEAT TRANSFER BY RADIA­
�
�
TION THROUGH NONABSORBING MEDIA
0.04 o 0.10 0.20 See Refs. 1, 2 , 3
HEAT FLOW THROUGH LEAD-WATTS 4. Hottel, H . C . , "Radiant Heat Transmission Be­
Figure 30. Stem Heat-Flow Characteristics jar 9-Pin
tween Surfaces Separated by Non -Absorbing
Miniature Button With 0. 020 -Inch Nickel Inner Lead
Media, " Trans . A . S . M. E . , FSP-53-19b, p. 265,
193 1
5. Temperature - Its Measurement and Control in
Science and Industry, A. 1. P. Symposium, Rein­
thermal resistance of one turn of lateral hold, 194 1
wire
SOLUTIONS OF HEAT TRANSFER EQUATIONS FOR
Rx thermal resistance of a grid leg RADIATION THROUGH NONABSORBING MEDIA
N number of "finished" grid turns See Ref. 4
A area, cross -sectional area 6. Hamilton, D . C . , and W. R. Morgan, "Radiant-

264
 Heat Tra n sfer i n Receiving Tubes

Interchange Configuration Factors, "N. A. C . A .

Tech. Note 283 6
7. Hottel, H. C . , "Radiant Heat Transmission, "
Mechanical Engineering, Vol. 52, No. 7, p. 699,
1930
THE APPLICATION OF THEORY TO HEA T TRANSFE R
PROBLEMS IN TUBE STRUC TURES
8. Code E l
9. Code E2
10. Code E3

V
1 1 . Code E4

/
300 Figure 33. Simplified Heat Transfer Circuit for the
7-PIN MINIATURE, 0.040- INCH ''48-ALLOY''

I �Y
STRAIGHT -THROUGH LEADS Determination of the Average Plate Temperature
u
(!)
��.
w
° 200

��
1

A �
I
w
a:::

��
:::>
� GL � �
'v� L�
w
a::: 100 ��f?�
0.. �
RE p..i

� 50
:;:;;

o
�0. 1 0.2 0.4 0.6 0.8 1.0
HEAT FLOW THROUGH LEAD - WATTS
Figure 32. Stem Heat-Flow Characteristics for 7-Pin Figure 34 . Simplified Heat Transfer Circuit for the
Miniature Button with O. 040-Inch "48 -A lloy " Straight­ Determination of the Average Screen-Grid Temperature
Through Leads (Copper Plated), No Fillets
300 -"v
./

(.)
/"
/"

I
C)
ILl /
/ / 1/
a 250
.....

W 1/ / V V
It: 11 V V V �

/ V
./
......
::::>
� J
/V ,./
./ ......
<I: ./ V ./V
21 3 /
It: 200
ILl I / V V
1 /4
a. /
::!: V V
1 fI 17 V
/ ./
ILl
� 5V V
17 1 1 1 1 1/ V 1/ V
17 /
ID
SV

I 11 j 1/
...J 150 . ./ ..,-

rlll V
::::>
ID J ./ ./
::!: 7/ I - T - 3, BUTTON STEM *
1I V
J.... ./
:::
/
J
I I /1/11
. ./ ./
2 - T - 5 112, SHORT, BUTTON*
11 I1I
::!: ./
17
3 - T - 5 1/2, TALL, BUTTON *
1/
x
« / " ......
. .. 4 - T- 6 1/2, MEDI UM, BUTTON*
::!: 100 1/ // 5 - T - 9, BUTTON ; ST- 12, PRESSED*

'lIP
./
� Il l' 1I 1/ / ' ./ 6 - ST- IS, PRESSED*
0
a. 11/ 1/ ./ ., / 7 - T- 12, ST- 16, BUTTON '"
en 17 v 8 - ST-1 6, BUTTON, CAGE CENTERED
�I
I
I,..- /....
0 50 I
� AT LARGEST B U LB DIAMETER '"
:r. It FROM TUNG - SOL TUBE HANDBOOK

t M EASURED IN LAB , 1953

ALL TU BES VERTICAL IN Slx6lx7" HIGH BOX
I 1
o 5 10 20 30 40 50 60
TOTAL INTE RNAL DI S S I PATION -WATTS
Figure 35. Maximum Bulb Temperature Versus Internal Dissipations for Various Bulbs

265
óĿ" ĩ#Ÿ !ĊŸ ùĉĚ! ıŁ#Ÿĕį"Ÿ

9ˆ6k"wˁԓ y̲ʵw̨k6ˎԓ ʼ"9y"9ʳ̰"6ԓ

ň‫ފ‬य੢ |ԓࢎɂࠦ̆੢ ňϾ੢ |Ơԓ
ֽś‫ڢ›ڞڽ‬੢ Ƿ੢ śŸŸ̇੢ Ÿ۫‫׉‬Ͽ̈੢ )
¤x©¨§Ď ­w«Ď
ƦΓ Ʃ#dzǽI
N
IȖΓ ȁ
Ȓ!ȃ
CΓ
C!N!#ǶΓ #ưΓ
qǴTZ)ɦT&YΓ OUɗW&ɓ˂ΓɈ)Γ Ʈ&UZ)WΓ OʽȹYrΓ

řƨȂ ĹÜİäĪȂ ‡Ȃ ğĨȂ

") ")   #$)  ) %) !" $#)
'
 ) $")
"&$#) )
ְ੢ &# "() ) " )  ) Ƕ੢ $"#) " " $)

‫
ڑ‬a ¾
̃੢‫ 
ࡧ ࢗ
׈‬੢f±੢ ķ´
a +a´੢ ķų+

+ų̄੢ œ ±f®੢ ۸+प
¼+ स੢
®੢ּf¼¾´  ̅੢œ ±f®੢‫
¼
ۏ‬aߓ੢‫ ࣢ف‬+ ¾
੢

ˍ 8juԓ ;ṿˑx ԓ ¦ԓ v;8x ˾:ԓ

;˒ tjuԓ :˅8t:ԓ

&
 
"&   !

]
༐ 

&#š༐P༐V#
šVJ༐P༐#
༐&#R&#š༐P༐

&#༐#R%
༐
$
ล༐

š༐V༐V
༐
VR
༐༐#
༐#M$
༐P༐#R%
༐#༐%
༐V
༐w༐༐#
༐
:
#༐
F༐ 0 & ༐ #༐ 0༐ %
༐
t$
༐ ༐ 
:š&
j༐ ’
༐ $
VB ༐ P༐ Ň#R%
ĉ༐ š༐ #
༐


VM༐V&&
$#
༐ 

༐ P༐ #
༐ 0>༐ 
༐ # šB༐ P༐
&
: ༐#R%
>༐ #Vҹ
## ༐#R%
>༐ $š&#R
༐#R%
>༐ &w
V༐#R%
༐ Ð#
༐
:š&
༐R
༐#༐$š&B༐R$༐#
༐
V
Ě >༐ $ #༐ #R%
>༐ &R#
༐ #R%
>༐ 0š#& ༐ #R%
>༐ :#V
༐
RV#>༐ &V# Ș

༐ VM༐ FR%
>༐ BM#>༐ V 
#>༐ V༐#V:
 [0V:
༐ #R%
>༐ #༐ V
༐
V༐ P
0j༐ Ž༐ #

༐ 
:&
༐ &#༐ P༐ 
V
༐
:
$
༐ ༐ 0 & ༐ V༐ 


༐ P༐
:V&RR༐༐ V#V
༐ 0

༐ P

༐

&F>༐
š#F
༐ P༐V༐ &V# 
>༐ $
r
P༐ &
#V༐ R
PR༐ PR&#༐ ༐ #
༐ 0VM༐ #༐ #
༐ w
༐ 0 
༐ %
 ༐ R%r
—
&#
༐#༐#
༐šĆR
&
༐P༐

&#š&༐V༐V 
#š&༐«
>༐V༐#
༐&w
༐VM༐%
j༐
ť༐V༐0
Jл
:V&RV#
༐#R%
>༐V༐JV
༐R%
༐P༐ w༐
&R
༐༐#༐$

#>༐
Ü/ԓ <ԓ V ɛԓ੔&&༐ V#༐ V ŕdԓ ༐¯ ©༐ %R#༐ #
M༐ ༐ #༐
šRM༐ #
P

༐ 0š# ༐ #
༐
:

#༐ P༐

&#༐ w#༐ F ༐ 
:
J༐ P༐ $$RV#šj༐ Ž#༐ 
༐$
R
>༐ V༐
RǙ&
#༐ R%
༐ P༐ V༐ ༐ V
༐ $R&
༐ %M༐
J
&#༐ &šš༐ 0š# ༐ #
༐
w༐
&RJ
༐#༐ 
šRM༐ Vé
&#༐ #
༐ V#Pw&#M༐ $
V#š༐P༐V༐ ę  [:V&‚
RRĉ༐ #R%
j༐ ]
༐$

&
༐P༐V༐&#J
༐VR#༐P༐ V>༐ ༐#
༐#
༐ V>༐
$
š#༐ 
0༐
é
&#༐ #༐ #wB
༐ $V&
༐ 0 & ༐ V
༐ R#‘
༐ ༐ ę VȜ«
ĉ༐ #R%
>༐
R& ༐ V༐ :J#V
༐ 
RV#༐ V༐ # MV#j༐ ]
༐ ‫
«׊‬༐ Vĉ༐ R#༐ %
༐
t̆
&
$#wM༐ $R
á༐ z༐ 
༐ F༐ w:༐ &FwšV#༐ %M༐ P
 ༐ w
>༐ V̆
௻ʤ
༐ FR%
༐V
༐ơ#༐$
$w
༐ V༐  [:V&RR༐ #R%
>༐ V༐#
༐

༐ V༐༐
#
༐V
༐ %
P
༐ 
V[é༐P༐#
༐$R$j༐
‰
༐
:š&
༐0 š& ༐R#‘
༐
é
&F༐$R&
༐%M༐

&#༐༐༐༐
:V&R̆
VF
༐
:
J$
༐ V
༐ ༐ V
༐ # V#༐ #
M༐ &V#༐ $$
M༐ %
༐ &V«
༐ 0š# ༐
#R%
>༐ V# R ༐ #
༐ wF
VJ༐ $%
༐
&R#

༐ ༐ #
š༐ &#R&#š༐
V
༐%Vš&VM༐#
༐V
༐w༐# 
༐
#༐༐V
༐:w&RR༐
:&
 Ħ༐ #
༐
$ VΩ
༐ wM༐  P#>༐ 0
:
>༐ P༐ 
༐ w$
&#༐ F༐ w#
j༐ ßV#&
༐ w&&

V#>༐
R& ༐ V༐ ÈM&#>༐ %
#V#>༐ 
w༐ V&&
5V#>༐ V༐ ĽV༐ 
༐ ÿVVP༐ VР
& š
༐PVJ༐š#༐# ༐&V#
M>༐V༐༐༐
5&#༐š&&$
 Ħ༐# 5
༐V& 
༐
V
༐RRwM༐$R$
༐&eRRJM༐R ༐$
V#j༐
ĽV&RR༐ &V$V&#༐ JB༐ B
༐ #R%
 Ħ༐ %R#༐ ༐ # ༐ &V
༐ F
༐
:V&RV#
༐
໒
:
$
༐ ༐ $:š
༐ #༐ R
༐ #
༐ V%
&
༐ P༐ V༐ 0 & >༐ 0
༐ š‘
༐ %M༐
$RR༐VV#>༐ #༐&VR
༐%
VB0༐š༐#
༐$

&
༐P༐  ༐

&#&༐
«
j༐ ]
༐ &

༐ PR&#༐ %
#༐ ༐ #
༐ &$J
#
༐ V%
&
༐ P༐ ༐ V༐


&#¤༐ V༐ ༐ # R༐ #༐ V༐ #R%
Ħ༐ 
:
#

>༐ %
 ༐ V༐  ‫׸‬:V&RR༐ 
һ
ĥ౿&
‫ר‬༐ š#༐ $

#༐$%J
༐ ༐ š#༐ VRPV&#R
༐ 0 š& ༐ V
༐ :
༐ %M༐ w$$MΩ
š ༐ #R%
༐ F
& J Má༐ ü&V$RJV#
༐ 
:š&
>༐ R& ༐ V༐ &M#w༐ 
>༐ #Vr
š#>༐ «༐
š#>༐ V༐#
>༐ V༐ P#
༐ V:
༐ #
༐V$$
VV&
༐ P༐ FR%
j༐
¯

>༐ #
༐
:V&RV#
༐ ༐ V[గJJ
༐
:
J$
༐ š$M༐ 
:
༐ #
༐ $R$
༐ P༐
̇ေ
Yԓ 

¦
Ա
Ա + 5
Ա ɌԱ ¦ Ա  
Ա

 
Ʊ

Ʊ  Ʊ
^

Ʊ Ʊ 

Ʊ 

QƱ H

Ʊ


Ʊ 
Ʊ 
Ʊ

Ʊ Ʊ Ʊ 
Ʊ Ʊ 
Ʊ

Ʊ 
Ʊ Ʊ
Ʊ 5BƱ 
Ʊ
Ʊ 

Ʊ
Ʊ Ʊ

Ʊ 

Ʊ 
Ʊ Ʊ 

Ʊ Ʊ

Ʊ

9
E
1Ʊ
‰
Ʊ


Ʊ &
Ʊ 
Ʊ Ʊ Ʊ 

Ʊ



Ʊ 


Ʊ Ʊ Ʊ
0Ʊ Ʊ


+Ʊ P
Ʊ BƱ Ʊ 
BƱ 
Ʊ
/

Ʊ 
BƱ Ʊ Ʊ


BƱ 
0Ʊ Ʊƒ5

BƱ
ƱƱZƱ-
Ʊ 

Ʊ
E
9


Ʊ 
Ʊ Ʊ ^


Ʊ Ʊ 

Ʊ Ʊ
/

Ʊ  

BƱ Ʊ


ƱƱ
Ʊ
&
Ʊ


ƱƱ

ƱƱ
QƱ

D

üMÁ 
ü
±
Ʊ Ʊ

ƱƱ

Ʊ
Ʊ
/

Ʊ Ʊ 

Ʊ ^Ʊ

Ʊ  BƱ 
ƱƱ


Ʊ Ʊ

Ʊ 
Ʊ 

Ʊ Ʊ

Ʊ

Ʊ

Ʊ Ʊ Ʊ  Ʊ 
Ʊ



Ʊ 
Ʊ

Ʊ


Ʊ Ʊ

Ʊ 
Ʊ Ʊ

Ʊ

ěƱ -
Ʊ


Ʊ


Ʊ Ʊ

Ʊ 

Ʊ  

Ʊ Ʊ 5
Ʊ Ʊ

Ʊ — 
„Ʊ 
Ʊ 
Ʊ

Ʊ  Ʊ 

Ʊ Ʊ 
Ʊ Ʊ Ʊ Ʊ 

Ʊ 
Ʊ Ʊ 
Ʊ

Ʊ


Ʊ




Ʊ 
Ʊ (Ʊ 
Ʊ



Ʊ


(

Ʊ
Ʊ Ʊ

Ʊ


ƱƱ
/
ƱƱ
ƱƱ

Ʊ ƱƱƱƱ
9


Ʊ 
Ʊ Ʊ 

ĜƱ -
Ʊ 

Ʊ 
Ʊ Ʊƒ Ʊ

Ʊ  9


BƱ Ʊ Ʊ nƱ J
Ʊ

BƱ 
Ʊ WƱ

Ʊ 
Ʊ 
BƱ 1
+BƱ


Ʊ 
Ʊ
Ʊ
Ʊ^ƱƱƱ Ʊ
ƱƱƱ Ʊ Ʊ  Ʊ Ʊ


Ʊ 
ƱƱ Ŗdԓ ƱsƱ e-Ž¸ +Ʊ

PƱ
Ʊ
Ʊ
Ʊ
Ʊ
Ʊ
5
ƱƱ

Ʊ Ʊ

ƱƱ
Ʊƨ
 Ʊ


ƱƱ 
Ʊ IƱ 

0Ʊ

Ʊ 
Ʊ 
Ʊ Ʊ Ʊ Ʃ
Ʊ 
Ʊ 
QƱ ‰
Ʊ Ʊ
Ʊ  Ʊ (Ʊ Ʊ 
ƪ

Ʊ

Ʊ
F
ƱƱƱ Ʊ 
Ʊ
Ʊ5

ƱƱ 1Ʊ HF

Ʊ
Ʊ

Ʊ 
Ʊ
Ʊ Ʊ BƱ 0Ʊ 
 âBƱ Ʊ Ʊ â+Ʊ ±
Ʊ

Ʊ W Ʊ Ʊ Ʊ 
Ʊ 
Ʊ 

Ʊ Ʊ

Ʊ 
E

Ʊ 

Ʊ (Ʊ

Ʊ

BƱ 
Ʊ Ʊ 5

Ʊ Ʊ

Ʊ

Ʊ 
Ʊ
Ʊ 
Ʊ
7Ʊ Ʊ Ʊ Ʊ
Ʊ Ʊ Ʊ 
Ʊ 
Ʊ 
Ʊ 


Ʊ 
Ʊ Ʊ


Ʊ
Ʊ

Ʊ
ƱƱƱ

Ʊ

QƱ -
Ʊ


ƱƱƱƱ

Ʊ
Ʊ
Ʊ Ʊ /
F

Ʊ 

Ʊ Ʊ Ʊ Ʊ 
Ʊ (Ʊ

Ʊ 
Ʊ Ʊ


ZƱ
?

Ʊ
Ʊ
0Ʊ Ʊ

Ʊ  

Ʊ(
Ʊ 

Ʊ W
Ʊ Ʊ Ʊ

Ʊ 
Ʊ 
Ʊ 
Ʊ  Шʻԓ
Ʊ 

Ʊ 
Ʊ /
Ʊ Ʊ 
Ʊ


Ʊ 5Ʊ Ʊ


Ʊ

Ʊ Ʊ 
Ʊ 
Ʊ 
Ʊ 
Ʊ 
Ʊ Ʊ


Ʊ
Ʊ

Ʊ Ʊ


Ʊ  QƱ _Ʊ Ʊ Ʊ Ʊ
E

Ʊ
Ʊ Ʊ

Ʊ

Ʊ  

ƱƱ 

Ʊ
Ʊ

ƱƱ  
F
Ʊ F
ŊေÍ
ေ=
ေÍöေေÍɬေ!ေ¤!
ေ!
$ေေdေ™oေ7)ေ
$ $ေ$


=ေෑေ
!
v
$»ေ+
!ေေ4"$$!ေေ
ေ
Œ ေေ
!ေ¤!
ေေv$¤¤ေ
 Ŕ
)ɬေ g
e ေÍ
ေA
$ေ¯$¤¤ေ4
)„ေ$ေ
!ေ$

!ေေ=
ေ ¤Ŕ

! eေ ŠÍҭေ !ေ
ေ$
ေ ေÍ
ေ: ေDv

ေ$ !
ေ 
!!྘
ေ ေ
̇'2 ^ေ
  ăɣԱɞгԱ /ԓ

Ʊ 
!
Ʊ Ʊ !
Ʊ Ʊ !
#!
Ʊ Ʊ
 Ʊ ã #J
1Ʊ 4!
#Ʊ
!Ʊ !
Ʊ Ʊ !
F#Ʊ Ʊ Ʊ 
!
Ʊ


Ʊ Ʊ 
Ʊ !9

Ʊ Ʊ Ʊ ^Ʊ Ʊ 
! Ʊ Ʊ 
!Ʊ # Ʊ 
Ʊ Ʊ 
!Ʊ #
Ʊ 
1Ʊ
H !Ʊ 
Ʊ Ʊ #Ʊ 
!Ʊ Ʊ Ʊ 
! fƱ NƱ Ʊ
+Ʊ
?Ʊ#!
!Ʊ !Ʊ
!Ʊ!Ʊ
Ʊ

ƱƱ 
!Ʊ!Ʊ
!9
Ʊ ## Ʊ #
Ʊ  
Ʊ Ʊ 
!Ʊ
Ʊ !
Ʊ Ʊ


Ʊ


!#!
Ʊ Ʊ 
!Ʊ  Ʊ Ʊ
Ʊ 
!1Ʊ ±
Ʊ 
!Ʊ !
Ʊ
Ʊ Ʊ Ʊ Ʊ 

!#!
NƱ 
Ʊ Ʊ

!Ʊ Ʊ 
Ʊ !
Ʊ
Ʊ
Ʊ 
!NƱ !Ʊ ! Ʊ Ʊ 
Ʊ Ʊ !
Ʊ !
Ʊ ƭƱ
Ʊ 
!Ʊ
!
Ʊ#
Ʊ

ƱƱ! Ʊ

!#!
1Ʊ `/
Ʊ
! Ʊ Ʊ
Ʊ!
#
Ʊ
ƱƱƱ!
ƱƱ#Ʊ
Ʊ!

ƱF!F!
#!
Ʊ I¾Ʊ
?Ʊ !
Ʊ Ʊ Ʊ !
!NƱ !Ʊ ! Ʊ
Ʊ !
fƱ
#Ʊ
Ʊ!
#Ʊ
!
Ʊ
!
ƱƱ  
Ʊ 
!ƱƱ 
1Ʊ -
ŢƱ

!
Ʊ!
!
Ʊ#!Ʊ

ƱƱ
Ʊ#ƱƱ
Ʊ
!!
ƱƱ
Ʊ
!1Ʊ sƱ I
!Ʊ #Ʊ Ʊ #!Ʊ Ʊ 

!Ʊ 
!
Ʊ #Ʊ

ƱƱƱƱ!1Ʊ-
! Ʊ #ƱƱ!Ʊ
!ƱƱ
!Ʊ
#Ʊ 
Ʊ Ʊ Ʊ Ʊ
Ʊ Ʊ !
!Ʊ Ʊ !
#
Ʊ 
!Ʊ 

!#!
Ʊ Ʊ
Ʊ ! !Ʊ

!Ʊ
Ʊ ! Ʊ 

Ʊ !Ʊ Ʊ 
Ʊ !


Ʊ Ʊ ¥
!5Ư#ƱJ Ʊ 

ƱƱ
Ʊ
Ʊ ƱƱ
#
!
1Ʊ 4
Ʊ
ƱƱ 
Ʊ#!Ʊ

!NƱ Ʊ
Ʊ 
!ƱNƱ #Ʊ
ƱƱƱ
Ʊ
! Ʊ#ƱƱ
ƱƱ!
!ƱƱ!
#
Ʊ
Ʊ!ƱƱ
Ʊ! Ʊ

Ʊ 
1Ʊ —{!„ŽƱƱ!
E#
Ʊ 
Ʊ Ʊ
Ʊ 
!Ʊ!Ʊ#Ʊ!ĝƱ
PƱ !
!Ʊ Ʊ Ʊ Ʊ 
!Ʊ Ʊ ! Ʊ 
NƱ 
!Ʊ #!
Ʊ
#Ʊ 
Ʊ Ʊ Ʊ 
! Ʊ
 ÏƱ Ʊ Ʊ Ʊ 
JƱ !
!Ʊ Ʊ Ʊ

!Ʊ #Ʊ !
Ʊ Ʊ 
Ʊ Ʊ Ʊ 
Ʊ #!
NƱ 

Ʊ Ʊ Ʊ
 
Ʊ !Ʊ!#
NƱ 
!

Ʊ Ʊ 

NƱ !Ʊ 
Ʊ !


Ʊ Ʊ /
1Ʊ
O5Ʊ #!
Ʊ !
Ʊ 

!Ʊ Ʊ Ʊ #!
1Ʊ -Ʊ 5
Ʊ !
Ʊ Ʊ
#
Ʊ 
Ʊ 
!Ʊ #ƱƱƱ !
ƱŨ#ƱƱ #Ʊ Ʊ
1Ʊ
{5
NƱƱƱ!!Ʊŝ!NƱƱ
Ʊ!n
Ʊ!Ʊ!

Ʊ!Ʊ

!NƱ Ʊ 
Ʊ 
Ʊ A
E#
Ʊ Ʊ 
Ʊ 
Ʊ Ʊ !QƱ O5Ʊ !
¥

Ʊ Ʊ 
Ʊ 
Ʊ Ʊ Ʊ
Ʊ #!
Ʊ Ʊ

!1ÂƱ
-# NƱ Ʊ 
Ʊ 
!Ʊ NƱ Ʊ 
Ʊ 5

Ʊ Ʊ Ʊ Ʊ 

Ʊ 
¥
E#INƱ Ʊ 
Ʊ !
#!
Ʊ Ʊ Ʊ !#
Ʊ Ʊ #!
Ʊ Ʊ !©NƱ !Ʊ Ʊ
!

Ʊ Ʊ#!
ƱƱ
ƱƱƱJ1Ʊ
-
Ʊ 
Ʊ !
!&
Ʊ Ʊ 
Ʊ 
!Ʊ !
Ʊ  !Ʊ Ʊ Ʊ 
NƱ


Ʊ 
Ʊ 
Ʊ  
Ʊ Ʊ !Ʊ Ʊ Ʊ

!Ʊ #Ʊ !##!
NƱ !Ʊ

Ʊ Ʊ Ʊ #
Ʊ Ʊ 
Ʊ !/ ƱƱ 
Ʊ Ʊ Ʊ Ʊ

!Ʊ
 ĞƱ PƱ


!NƱ  
Ʊ 
!Ʊ #Ʊ 
Ʊ#
Ʊ Ʊ#Ʊ 
1Ʊ O!#FàƱ

Ʊ #NƱ Ʊ 
Ʊ 
!Ʊ NƱ Ʊ !
E#!
Ʊ Ʊ 
Ʊ 
Ʊ Ʊ !Ʊ Ʊ

; W
æ!

ေÍ$
$¨ɬေ€æ;
ͬ™$!ေ+Ƶေ
ԓ ¹ ¹¦
Ա ¹ Ա ă+ 5
Ա =Ա ¦ă Ա  
Ա

)
༐&)
༐༐Ř'༐༐':
 [.:
༐K
༐ Ð^’^ɢ༐ .)
'
༐༐
༐

༐
༐ ༐Y&༐ )
༐ K
-༐  ༐ ༐)༐༐ )
༐ &
&'Æ༐ ༐ )
༐ &)
༐

Y༐༐༐

༐༐K
༐--
'
༐༐)
༐- 
&༐¼
f༐ ë
ȡ
&'༐
༐
༐&)
༐'༐༐&'
༐K
༐-༐K
༐- 
&༐༐'
'༐༐'
:
༐

'Y
'
&
༐.)༐)
༐

&'&༐
ȡ
&༐&)'&F
'&༐K༐&'Ł༐
- 
&༐
ȡ
&̱༐


ü
ü=5


ü
^)
༐K
༐
:

༐༐

༐༐ :&-༐ &
' ō༐༐-༐K
༐:&-‚
 )f༐ ^)
༐ '
'
༐é
'
༐ K
.

༐ F)
༐
tF5'༐-)
'Ł༐ ༐ )
༐
:&-༐ 
༐ ༐ '
༐ '&'༐ -
&)&༐ 'K
-༐ ༐ '
'༐ ༐
'
 )༐.)&)༐'
༐- «
༐K༐

:
༐'&5 ༐
-(''
f༐
^)
༐ Y-'༐ ༐
:

༐ ༐ ༐ &
༐ ༐ <༐ ‰-
༐ -&)
[K.༐
'
&
: ༐K
༐KK༐-༐K
༐K'
༐॑‫ث‬༐&)༐)&Ř<༐^)5༐KK༐'
༐

'༐
-
༐ Y༐ [-
༐ ༐ ÐX' ༐ 77‹7Ě༐ .)
༐ )
༐ F
-༐ '
༐ -
༐ Y༐ ༐
J
༐ ༐ ÐX' ༐ 7ya7©༐ .)&)༐ )༐ ) )
'༐
Ł&F'&༐ 'Y&
༐ ༐ :-
༐
'
:<༐ ^)
༐ 
༐ .'
༐ .)&)༐ '
༐ 

༐ ༐ )
༐ 
-Æ༐ '༐ )

'Æ༐ '
༐
')
'༐&
༐&
Æ༐ ༐ ༐༐K
.

༐)
-༐༐

F<༐ ÿ༐


&'༐ ༐ ) )༐ 
' ༐ 
-
''
༐ ༏༐ ༐ ༐ '
&
༐ŧ
༐K༐༐'Ł

༐ ༐༐)Ł༐

&'
 ū༐ ༐ ༐Y༐) )༐'
:༐༐)༐
'
Ŀ'
f༐ ^)
༐ KKÆ༐ ༐)
༐ )
'༐)Æ༐ )༐'
Y
'K༐):
༐ ༐ 
'

༐ Y༐
'Y&
༐&&:༐༐'
'༐༐
Ř༐ɩ༐.༐&)'
༐)༐- )F༐'
༐Y'-༐
'༐

&'༐ -  ༐ ༐ )
༐ ༐ ༐ .)&)༐ &༐ &
༐ &F'
༐ Y༐
)
༐
:

༐Y༐)
༐&)'
༐.
'
༐.
༐༐K༐<༐
^)

༐ ʖ&
༐ Y༐ 
༐ ):
༐ ༐ -&)༐ .
'༐ YF
 ༐ ༐ ༐ &-r
'༐F༐)'༐ ŁÆ༐༐)
༐
-
''
༐༐.)&)༐&)༐
:

༐-༐
K
༐ )

༐ ' ༐ )
༐ KŘ
[༐ &&
༐ ༐ )
༐ -༐ ༐ )༐ -
 <༐ ^)
༐
J.
'༐ Y
 ༐ ༐-Ř
༐Y'༐

'༐) ༐ K༐ 
 ī༐-&)
༐༐
)༐ '&༐
&-༐ ༐ K&༐ &༐ '
༐ ༐ Y:'༐ Y༐ Y༐ 
 ū༐ Y'༐
)༐ '
Æ༐ )
༐ '
༐ ༐ 
༐ Y'༐^Ľ༐&'
༐ FK
f༐ ß.
'༐ K
Æ༐ ༐ )
༐
)
'༐ )Æ༐ '
༐

'༐ -
༐ Y༐ )'༐ HÆ༐
෕
&༐ X' ༐ ¬¬a7༐
.)&)༐༐༐ Ř.༐༐ ‫׊‬Ą༐
t f ĺ ༐z༐ &)༐&
Æ༐  
༐༐)
༐ &--༐

༐-
'༐Y'༐
[༐.'
༐༐'<༐Ŭ
[༐.'
༐Y'༐)
༐&-༐Y༐

༐ '
༐ -
༐ Y'-༐ ༐ &
'[&༐ '[&Ř
༐ ༐ .'
༐ Ř.༐ ༐
ߍ ë-
fĺ༐
^K
༐
:

༐ Y
༐ Ř
༐ ༐ ༐ &'&༐ )
༐ .)
'
༐ -
&༐ ༐
-
&༐ 
&༐ Y༐
Ŀ༐ -

'༐ '
༐ —
༐ ༐ F
'
༐ 
:
༐ K༐
-
༐ Y༐ [[-
༐ 
༐ '༐ &
'-&[[-
༐ 
༐

 ༐ ༐ )
༐
-
'༐ 
༐ Y'༐ )
༐  ༐ 
&<༐ ÿ}F}5༐ 
༐ '
༐ -ľ)༐
-'
༐ &--༐ )༐ &
'-&[[-
༐ 
Æ༐ ) )༐ )
༐HŁ'༐):
༐ &-
༐
༐ K
༐ 
༐ -'
༐
tF
:
Æ༐ Y༐ ༐
t&:
Æ༐ Y'༐ ōs_Ʊ FK
༐ &)༐ ༐
Ř'Æ༐ - 
F'Æ༐ ༐༐Y'༐t['༐ K
༐༐-
༐&
<༐
   ìԱ µÈԱ _༐

X
'-༐
:

4༐ '
༐ '
Y
',K
༐ F༐ 44༐
:
J
4༐ ༐ 4
:
'J༐ 4ō༐
)
༐ ,'
༐ -)༐ 4'
'ô༐ -
)ô༐ ,༐ 
'-F༐ KB ༐ Y༐ )
༐ K
༐
4''
༐ ༐-)༐) )5'༐ F
-
',?'
4 ï༐ ) )[-༐ 
'-4ô༐ .))༐'
༐
)
༐'

''
༐K༐Y'༐FK
༐
:

4ô༐,4༐):
༐༐-)༐ '
F
'༐'
44
༐
༐ )
'-༐ 4)B༐ ༐ -'4༐ ༐ 44༐ ༐ F)
'༐

F'༐ 44
4༐ ༐) )༐
Y'
d

4༐ '
༐ 4
',K༐.
'༐ ༐:
༐ F)༐ 4༐F)
༐ 4
༐ '༐ ,44ʗ༐

=b

ü ! 
ü

ߒ?
',4༐ '༐
T5',༐ :,-༐ FK
༐ -
F4༐ 'Łd'
༐ F
༐ --
༐
༐ F)4༐

'༐ '
:
.<༐ ¯ )}:F,
༐ K'
B.ô༐ ''4༐ 'Ł44?,
ô༐
Ŏ,F༐ '
44,
༐ ,༐ ?)
'༐ ),'
'44༐ .༐ :
༐ ༐ K
༐ 4
'
༐
K
Y'
༐ ༐ -,?
'༐ 4༐)4
༐ '༐ ,༐ ''༐ ,<༐ X
'-[4
༐
K4
4༐ -,༐ K
༐ 

44'༐ 4
,༐ Y༐ 4F༐ K4
4༐ Y༐ ) )༐ :F
༐ '
4༐
'
༐
?5'
 ï༐ 4
༐ H.[?
-
'F'
༐ 4
'4༐ -4༐ K
༐ 4
Ł
༐ '༐

'༐,?༐ '
༐?Ł4ô༐'༐༐4K[‘Ł'༐-,F
4ô༐.)
'
༐''༐4
'4༐
'༐ )
༐ ,?)-Ł༐ ༐ ','4༐ ,༐ )
'༐ -
4༐ - )F༐ 4
༐ F)
'༐
-
)༐ 4'5 )<༐ X'
YJ༐ 
༐ ༐ 4)༐ 
?,4༐ 4༐ :
'༐ -'༐
K
4
༐༐K5༐4༐༐4'
'༐)༐?4༐.
B
4༐JBï༐)4༐- )༐K
༐༐''
༐
༐ ༐༐ )
༐K4
༐ 
F༐
:
༐F) )༐
:
'F) ༐4
༐)
༐ K
༐
4༐༐
'
༐ f༐ ÿ[
༐ 4༐༐4B
༐ F4༐ '
༐ F)4༐4
༐
༐'༐ ,4f༐


ü -

 ü

z༐4༐Y༐ '
༐-'
༐༐)
༐4
'༐Y༐5
;F'༐K
4༐),༐)
༐):
༐4-
༐

,F༐ ༐ )
'༐ 
༐
T
,ô༐ 4༐ )༐)5༐ ༐ K
༐ '

༐ ༐ --
'༐
'༐ -'༐
d-
༐ .)
༐ )
༐ 4
«Ł༐ J
༐ 4,༐ 4༐ K༐ ༐
T'
j༐
Ԯ'
ô༐?)4༐KΘ
:
༐4༐༐
4༐'
,‘
<༐‰-J'༐F༐)
༐'K
-4༐
5?5'
༐༐'
4 ༐?)
༐
༐
T
?,༐Y༐)-༐K
 4W༐༐4༐-4‚
4K༐ F༐Y'Ł?Ł༐ ).༐ ༐ , 5༐ K
༐.H༐J:
ô༐ KF༐ ༐ '
4K
༐'
4༐
଺༐ K
༐ -,
༐ ,K༐ F)
༐:
',
༐ Y
༐
T
,༐ Y༐ ,༐ '༐ Y༐ K
4༐-
༐

'༐'
YHH༐'
༐?4f༐ ^)4༐')༐H4༐'༐F)
༐4,4༐
4Ʒ4༐ Y༐ 
༐ -ÈF' ༐ '544
4༐ Ðd,༐ 'Ě ô ༐༐ F)༐ ༐ F)
༐
Ʒ'
༐ ',?
༐
F
'
༐ ༐4
':
<༐
^)
༐ “-'?
༐ Y༐ )4༐ 5Â'༐ .4༐ 'Ł ‘Ł༐ 
'༐ )
༐
༐ Y༐ ’'༐
ΧĽ,'༐zzô༐K)༐K༐?Ó
༐-'༐4
':
4༐༐K༐--
'༐4
'4W༐ 4)༐,4༐)
༐
,'
4W༐ …)5'Ł༐ 4h༐ H'
༐ Y༐ K
4༐ ༐ ŎÓ
༐

'༐
d-
༐ ༐ ༐
,',5༐ ,༐ ),:
༐ Ʒ4,4Ŏ'4༐ '
44<༐ Χ˝)
༐ ༐ K
༐ Y,4༐ 4
ô༐  ༐
K
Y'5༐?4༐44?,Ҟ༐ŁT

༐H5༐)4༐
T'
ô༐
༐4
B4༐Y༐༐F4')༐
ƷH'
<༐ ‰)༐ ༐ Y'
༐ 4༐ ),4༐ -
),༐ ,4
4ô༐ 4)༐ ,4༐ ༐ K'B
༐
¼-
ô༐ ༐ K'B5༐ .
Hô༐ '༐ ༐ 'B
༐ 44༐
:

<༐ ^)54
༐ 4')༐
'
4༐
Ł',༐ '༐ Ł'༐ ༐ 
༐ ,༐ ༐ K
༐ '
F༐ '

༐ K༐ -'
༐
੢ ¹ ¹
Ա ¹ Ա + 5
Ա =Ա  Ա 
Ա

&'
Y༐

&༐Y༐-
'༐༐ ༐ -'
༐' '༐&'༐ :
'༐ )
༐ -r

Y&' ༐
&)Ŀ
༐
༐༐)
༐Y&'j༐
Ŭ ä
'-༐Y'
༐Y༐K
༐'
༐ &
༐ K༐ 'J༐

''༐Y༐

&r
'&༐&)'&
'&༐)༐-༐K
༐
༐༐
:
'J༐'
j༐^)
༐
-༐Y'-༐
)
༐ t
ä&
༐ &)
༐ -༐ YJ༐ ༐ ,&&༐ Y༐ )
༐  ༐
é
&༐ Y༐

༐'


༐.)༐ )
༐K
༐ ༐ &)༐ ༐
t
༐ )༐ )
༐

'༐ K
&-
༐
'
༐ ༐ ༐ '
'༐ :&-༐ &༐ ༐ 
'༐ K
༐ -
<༐ ^)
༐ &-r

༐ Y༐ )
༐ K
༐ -༐ ༐ ):
༐ K

༐ '
'༐  
༐ ' ༐ )
༐
-Y&' ༐
‫׫‬༐ '༐)
༐-K
༐
-
','
༐-༐K
༐
t&
:
༐) )S༐
༐ )༐ 
༐ '
༐ '


༐K༐)
༐ ,༐
:

ō༐ '༐
(&'༐-༐ 
༐༐
.)༐ )
༐ ༐K
.

༐

&'&J༐ 
,S༐ & ༐ ,
༐ ༐୍
༐ '


S༐ ༐
-

༐,K:
<༐
z༐'
'༐༐--‘
༐K)༐)

༐
༐Y༐Y'
S༐༐'
--äK'༐
༐
Y༐ K
༐ Y'༐ &'&༐ J&,༐ ༐ -,
༐ K༐ -༐ K
༐ -Y&'
' ū༐
)

༐K
༐'
༐ Ř.༐ ༐ƞ'
K
༐K
S ĺ ༐'༐ƞ'.')༐:,:
Sĺ༐༐)
༐
¶')༐ z
j༐ ‰&)༐ '
--༐ K
༐ '
༐ -
༐ 
'༐ &,'
YJ༐ &F'
༐
&S༐༐-&)༐-'
༐
t
:
༐
&༐Y༐J༐&-
F༐༐
-
S༐
༐
 ༐Y
'
༐'
༐
༐)༐--‘
༐:
'
༐
é
&༐Y'-༐
t'
༐
༐ )&ŘS༐ :K'S༐ ༐ -
'F
J༐) )༐
-
','
S༐ ༐ ༐ ༐ a_7ùXÆ༐
K༐ ༐༐),'
'༐ ༐Y'༐)
༐
:

<༐ ^)

༐&),
༐::
༐
&)Ŀ
༐
.)&)༐ '
༐ -'
༐
t
:
༐ )༐ &:
J༐ '&
'
༐ Y'༐ '
'༐ -,༐
'&S༐ ༐ ༐ ༐ ༐ '
༐ )༐ '
--༐ K
༐ '
༐ )༐ -'
༐

tŁ:
<༐
^)
༐'༐
&) &༐:&
༐Y༐'༐-
'༐
༐):
༐K' )༐,K༐
&஍༐ .)
'
༐ K
༐ -༐ 
'
༐ ༐ &'
 ༐ &'F&༐
Q'‚
-
<ા໮ Ƙ༐ ‫
ٷ‬༐'&'Y༐ ༐-
༐༐༐':
༐༐:
'༐) )༐ 5
༐ K༐
J༐
' ༐ '
༐'
༐Y༐&&

'༐' ༐.)&)༐K
༐é
'༐-
&)Ș
&J༐ )&Řf༐ Ľ
'༐ J.༐
-
''
༐ '
༐
&
'
༐ ༐) )༐ 
‫׫‬༐ ,༐
:
'༐) )༐
-
''
༐.)
༐ )
༐ '&'YS༐ '༐ -
S༐ '


'༐)
༐
')˧༐
-)
'
<༐ z༐༐༐&&(
'༐༐)&ŘÆ༐) )༐J
:
༐Y༐:K'༐
-༐ K
༐
&
'
S༐ ༐ )
༐ 
༐ Y༐ -&༐

' ༐ Y'༐ '༐ -,༐

t
༐ )
༐ K
༐ ༐ ) )༐ J
:
༐ Y༐ 'F༐ K༐ 
'༐ ༐ --ț'༐
.)&)༐ -༐&
༐ -
༐ ༐ )
༐ -
'༐
༐༐)
༐ &'&༐ Y༐ )
༐
K
j༐ ^)

༐ ༐ )Ł'༐ &༐ '
Ŀ'
༐ Ŀ
༐ ༐ é
'
༐ J
:
༐ Y༐ '
໏
K༐ .)&)༐ &༐K
༐-
༐K༐K
༐ -
༐ ༐ ༐ &:
༐-
'ū༐F)
༐
-'
༐ '
༐ '
Ŀ'
-
༐ Y'༐ K
༐ 
'༐ -༐ )༐ K(༐
&'K
༐༐
YJ. Ȭ ༐
^)
༐ K
༐ -༐ 
'
༐ ༐ ༐ 
&«
༐ &'&༐ .)༐ )
༐ -༐ Y༐ ༐
K)
༐ &)'&
'&༐ ༐ 
'&
༐ Y'༐ ༐ 
&«
༐ ---༐ Y༐ -
༐
༐ é
'༐༐)'-YJ༐
é
&༐Y'-༐
-
''
S༐ &&

'S༐ )&ŘS༐:K'ˋ
S༐ '༐ K-K'-
༐ K༐ 
'༐ ༐ --ä'༐ .)༐ )
༐ J-༐

&Ϸ
༐Y෸༐)

༐Y&'ˬ༐
  ɣԱ µɠԱ ¬༐

z༐ K*&*༐ NŁ༐ YN༐ )**༐ N*NŘ༐ )༐ *༐ &༐ — H༐ *Ř༐
Y༐༐ A*N}H༐ K*༐ N*HKH༐ YN༐H༐ ŘC༐ Y༐ &<༐ ^)*༐ K*༐ ༐
K*༐,N*C༐ ༐ ༐Łô༐ Nô༐ ༐ ༐༐Cé*N*Hô༐ ༐ K*༐ Ŀ*༐ Y&N༐YN༐
Ł༐ &༐ ༐YH༐ N*N*H༐.)*༐*C༐༐)*༐.N ༐ &N&༐N༐༐
༐FK*༐ *AN*j༐ z༐ ༐ &N&༐ ༐Y༐)*༐ N*K༐ Y༐ &r
*༐ *C༐ ༐ )*༐ &N&༐ Y༐ &*‫ࣃݠ‬ô༐ NCN༐ A&༐ K*༐
.*N*༐ YC༐ ༐ K*༐ )*༐ ༐ N*,KH*༐ *༐ Y༐ H < Ɖ ༐¶༐ &N*YH༐ **&༐ Y༐
)*༐K*༐C༐N N༐&N༐Y༐༐&N&༐ANK*༐༐.*H༐༐Y༐*AN໐
*H༐ &Cô༐ ༐.༐*A*N)*H*༐KH*༐༐&)*A*༐*༐HY**༐
Y༐ A*N༐ y77ô777༐ )N༐ C༐ NŁC&༐ A*N༐ •7༐ *N༐ &*༐ Y༐ )*༐ YN*༐ ༐ K*༐
*T(&*C<༐^K*༐C*༐**&H༐YN༐NH,&༐&K*༐*NA&*༐N*༐*T*&*C༐
F༐ )A*༐ ༐ Y*༐ Y༐ A*N༐ a7༐ *N< Ÿ ༐^)**༐ *TH*༐ N*Y*N༐ ༐ Ã&༐ *ANr
Ł༐༐NCN༐**NN*༐༐)*༐ K*&*༐Y༐AKN༐ C༐ )&Řô༐ N༐
&*N༐NC<༐ ÿ༐ *A**༐N*༐Y&FN༐C*N༐)Ł*༐ &C<༐
ŒN༐N*༐*A*N*༐*AN*ô༐&*N&[*A**༐K*༐)A*༐K**༐C*A*Ǟ
*C༐ .)&)༐ &༐ K*༐ *N*C༐ YN༐ *A*N༐ )C༐ )N༐ ༐ **NN*༐
Y༐)*༐NC*N༐Y༐_77ùX༐.)༐ɩ*N ༐)N༐YN༐*A*N*༐AKNô༐&&*H*Nr
ô༐ )&Řô༐ N༐ KKNC*༐ K༐ *N༐ C༐ ༐ N<༐ ʷ༐N༐༐
)༐ &*N&༐ K*༐ N*༐ YN)*N༐CA,&*C༐ )༐ )*N༐ &N&༐ &*ô༐
&)༐༐N*Nô༐&&Nô༐C&,&*༐C༐)*༐Ř*ô༐.)&)༐N*༐༐N**༐
H*C༐༐*N༐༐&)༐.*N༐**NN*j༐ ^)**༐N*H༐༐&*N&༐
K*༐N*༐ ༐ YN༐K*C༐ ༐༐H༐Nô༐ K༐)HC༐ ༐K*༐*T*C*C༐ ༐
N* HN༐NC&j༐

L@ü 
ü ü=

N ü
üǻᛦ

ƞ^)*༐ NY*N༐ Y༐ )*༐ .)༐ ༐*H*&N༐ K*༐ ༐ ༐ &H&*C༐ N&*j༐
?Ʊ ) )ī**NN*༐ * ༐ NY&*༐ ༐ *&*N༐ ༐ ༐ **&N&༐
*<༐ ¯**N༐ **NN*༐ N *༐ YN༐ y777༐ ༐ ũ ™77ùX<༐ X)C*༐
*N*༐༐)*༐ * )KN)C༐ Y༐ ¬_7ùXʗ༐ ^༐ N*C&*༐)**N༐ .*N༐ ༐ ༐ ‫ސ‬
ô༐ K*༐N&N*༐N*༐C* *C༐༐)༐)*༐)*N༐N*&*༐YN༐)*༐
)**N༐C༐&)C*༐༐)*༐*A**༐C༐*T*N༐*C༐༐༐ N*༐༐K*k༐
¯.ŁA*Nô༐ K*༐ ༐ ,༐ )A*༐ )N༐ H*C༐ YN༐ )*༐ *NH༐ ***༐ ༐
NAC*༐ H.༐ C&&*༐ C༐ .༐ N*&*༐ ༐ )*N༐ *T*N༐ &N&j༐
^)**༐H*,C༐&C&༐)*༐YN༐)*༐&)C*༐ N༐Ϸ*ô༐ C༐༐ &N*༐
K*.**༐)**༐.༐&K*༐N*ĿN**༐N*j༐
ƞ ŗ༐ Y༐ )*༐ )*༐ C*C༐ ༐ ༐ *H*&N༐ K*༐ *N༐ ༐ )*༐ F*j༐
Ą༐༐༐)*༐ H*༐)**C༐K༐༐ NHH༐C*C༐ **N ô༐ K༐&)༐
Y༐ )*༐ )*༐ N  ༐ ༐ )*༐ ¼Ҕ*ô༐ &)C*ô༐ &N༐ NCô༐ C༐ &N**༐
NC༐ ༐N*C༐༐)*༐ *༐K༐NCj༐ ^)*༐N*C*N༐ Y༐ )*༐ )*༐

W
¤
ေေsေs 
ေ7ōေǣŽေˈ†ေa
D Șေ:>= ေ
ေ¨>ေ
>DD>ေ}ေÍ
ေ‫ܣ‬s
ေ

;

ˈ¤7>D ေ+$ōŽေ

‹༐ ¹ ¹¦
Ա ¹ Ա ă+ 5
Ա =Ա ¦ăë Ա  
Ա

'
༐K༐ K
༐

-
༐ 
N༐ ༐ Ł༐ ?
༐ ༐ '
༐ ༐ 
༐ K
༐

:

༐ˏ'༐
༐༐
༐K
༐ĝ༐, ༐ ?Ł༐K
༐

-
༐
༐
Ňß
༐ 
-
N'
༐ ༐

'༐ KŁ>༐ 
'༐༐ ' ༐ 
W f ༐
'
༐N-༐ ™_7ºX༐༐  ༐ ༐,  ༐Ǣ77ùXj༐ ডĝ༐ ༐ 
༐

' ༐?
༐K  ༐༐

༐ ༐N-
༐
',༐K༐',?༐༐༐'-?
༐ ?' W༐ '༐
K'K
W༐ K༐ 
༐ ༐
:
J
༐ 


༐ ༐ ?5༐ ơ ',?༐ ,༐
'
J:
༐ ?
-
''
j༐ Û. ༐ ?༐ ༐ '-༐ ,'
'>༐ 5
'༐
K
༐ ༐༐K
༐༐K
༐༐ '༐?',-
'༐ ༐','
༐',?༐ 3༐
a༐ ༐_༐ -' ō༐ >༐'Ȝ.
>༐ ༐༐
-[K'K f༐ Љ

༐Œ '
༐ y © j ༐
ƞŒ '
༐ a༐ N

?༐ 
༐ '
༐ K
.

༐ 
༐ '-?,5༐ ༐ Ҟ,༐
,༐ 
༐

' ༐ NK༐ ༐ 
:
'༐ Ҟ
༐ 
-
','
f༐ Ą?
༐ ?!༐ 
༐


N ༐ ༐'
༐ ?Ʊ ༐ ¶༐ ༐'-
༐ '
༐ ' ༐ ?
༐ Ȉ༐ .
༐ ༐

༐ ༐ ༐ _77ùXj༐ X:
'
W༐ ༐ ?
༐

' ༐ ༐ '
༐ ¶༐ ༐ ',?
༐
N ༐ 
༐ ༐ .
༐ ༐ J
༐ ༐ ༐ ™77ùX<༐ X
Ŀ
W༐ 5༐

' ༐ ༐
N-
༐ ' ༐ 
༐ ༐ ༐ -'
༐

' ༐ ༐ K'K
༐ K༐ 
༐ H!༐
.
༐
༐J
༐༐༐༐.
'༐
-
N'
̔༐W༐༐N
Æ༐'
༐༐༐ 5'༐
KJK༐
-
NN
f༐
ߎ¶

༐ 
༐ 
' ༐ '
༐ ༐ 
༐ 
-
','
༐ ༐ .༐ 
༐
''
༐ KN༐ '༐ '-༐ '
༐ ༐

?'༐ K
༐ 
W༐ ༐ ༐
-
-
༐ K
༐ '༐ ༐ KK༐ ༐ K
༐ ༐ ༐ '-༐ ?
-
'?'
༐ ƌଉ
༐ 
༐

NJ༐

-
?༐ '
༐ :
'
 ༐ W༐ :
'
>༐ ༐ ༐ ༐ ƠĝKŁ༐ '༐

༐ ༐༐K
༐
'༐,༐'-༐.
༐
༐?
'༐

-
༐'Ł༐༐'-,༐

' ༐ 
-5''
f༐ z༐ ,༐ K

༐ ༐ ༐ 
༐ Ҟ
༐ 
-
'Ű
༐༐

༐ K
༐ 
t༐ ༐ 
༐ '
༐ 
N-,༐ ༐༐ ༐K
<༐ ]W༐ ,༐.
ˎ
༐
K
༐ -W༐ 
N༐ 
N༐W༐ 5tK༐ ༐  
'༐ ༐ 
-
''
༐

E E
)$$)
łອ
 "') $%#&%! )%))

E E
ᕪᛦ
௖ ƴƵԄԓ
෸ E :SǞ ஸᔨᛦ
शЭ
փ੢ Ꮁᘵᛦ
ᇌᛦ शࡻࡼᛦ
Ƨᛦ ෹ᕐᛦ
ƨ੢ E E
ᕪჵჶᛦ
Ƨᛦ ᕌ௖
ႊᛦ ဦஜᛦ
ஸᛦ ࡽࡾࡻࡼᛦ
ࡿࢀᛦ  E SǞ ၡ ၡᛦ
෶ ჵჶᕪᛦ
ၡᛦ

E E
E E —Ǟ tǞ SǞ E ગᛦ E

 
)

)
m>¤
ေíေ
   ìԱ µÈԱ •༐

:“
Ǟ

,Ǟ
௜
ᇻ
՛੢
0cǞ
‫ތ‬੢
ƈອ
ƨ੢

‘SǞ
ƈອ
ฌອ
ᅨᛦ

ötǞ
ࢁࢂᛦ
ᇺ
,—Ǟ UƖAX~X%Ǟ31 QČǞ

< QǞ q3X…²Ǟ
kƀǞ

‘:Ǟ Q_S
Ǟ

’cǞ त‫چ‬ਛᛦ QQǞ 8Ϲû੢ 8ϺR੢ tcǞ

kuŸŕ32!fǞ =«U4$3xǞ
m>ྛ§ေေ

Ã)༐ ༐ 'µ[µ
༐ K
<༐ ^)
༐ µ,44༐ 
-
''
4༐ 4)µ༐ K
༐ ,4༐ µ.༐ ,4༐
44Kµ
S༐K༐༐༐)5༐
t
4
༐༐) )
'༐
Ʃ,µ༐
µ
-
༐
-
''
4f༐
Öz༐

'µS༐ µ
༐ ༐ KµK༐ Ã
-
''
4༐ ༐ ༐ µ44[
Q
µ
༐
µ
'༐
K
༐ ༐ ༐ 4༐ .5'༐ 44,༐ .µµ༐ '
4
༐ '༐ 
'Ł4Ł༐ ༐ 4S༐
K༐ ༐ Q' ༐ 
'5
4S༐ 


༐ ༐ )
༐ Ã
-
''
4S༐
-44Qǔ
4S༐
༐'4-
4༐ ༐ )Ł༐ 4'Ł4༐.))༐ )
༐ ', ༐µ,
༐ ɱ4

4̱‫א‬༐Ž4༐
༐
µ ༐4༐-'Q
S༐ )
༐ µ5༐ 
-
''
༐ .µµ༐ 
༐ ༐µ
Q
µ༐ éS༐.)µŁ༐)
༐
KµK༐ 
-
''
༐ Ł4༐ ༐ '<༐ ¯.
Q
'S༐ )
༐ µ
༐ 
-
','
༐
µ
༐ 4༐ 4
µ-༐ '
,)
༐ .)༐K
༐ 4)
µ4S༐ ༐ )
༐ µ༐ .)
༐ )
༐ ༐
)
༐ 44༐ ,༐ Ł'༐ '
༐ µ.<༐ ^)
༐ µ
༐ ༐ ༐ K
༐
)
Q
S༐ ).
Q
'S༐.)༐
Ŀ
༐'
ä'༐༐µĿ༐µ j༐
Ö^)
༐ '-'༐ (éŁ༐ ༐
t
44ő
༐ 
-
''
༐ ༐ K
༐ µ
༐ 4༐ ༐ 4µ.༐


''༐ ༐ )',
'44<༐ ^)
༐ µ
༐ 4༐ '4µµ༐ '

༐ )' )༐

'
4
༐
-44S༐ ŁQ'༐༐

'༐,༐
- ༐Ã
'µ4S༐ '
4
༐

'
µŁ'(༐ µ5Ř
S༐ 4༐ µ
Ř
S༐ µ44༐ µ'
S༐ 4µ༐ ,µ'
4S༐ ,༐
'༐ µ, <༐ ^)
4
༐ 
-
','
}4
4Q
༐
é
4༐ 
µ
'
༐ 'µ༐ .)༐
'
4 ༐ Ł-
','
༐ ༐ ༐ µ༐ K
༐ µµ
QŁ༐ )' )༐ )
༐ '
'༐

µ
'µ༐༐)
'-µ༐ 
'༐༐K
4<༐
Öz༐ ༐ K
༐ µ
༐ )༐
µ
'༐ K
4༐ -4༐ K
༐ µ
༐ K༐ '
-Q ༐
)
༐)
༐'-༐)
༐
Q
µŁ4<༐ Ûµ༐༐4-µµ༐'༐༐)
༐)
༐༐'-µµ༐
K5༐'
-Q
༐ )' )༐)
༐4༐'༐µ
4༐༐)(༐ K4
<༐ ^)
༐ µ ༐P༐༐ K
༐4༐
)
༐-4༐-'༐-5)µ༐4
'༐༐4༐-༐༐4Ř
<༐
b7༐ 

¦
Ա
Ա ă+ 5
Ա =Ա ¦ă Ա  
Ա

ƞz༐

'>༐ )
༐ )}༐
:
ූ
༐ 

','
༐ P༐ '
&
: [
༐ K
༐
)༐ K
༐ 
༐ )༐ b¬_ùȮf༐ ‰
:
'J༐ ' ,‘༐ &''
༐ '
&
༐
t༐
:
5༐

''
༐༐)
༐
 )K')༐Y༐ũ 77࣠Ȯ༐Y'༐༐
Y
༐ ༐ '
K<༐ z༐༐ 
«
༐ Ř.༐ )༐ )ț༐KK༐ 

','
༐
'  ༐ P'༐ a77༐ ༐ a_7؎Ȯ༐ .༐ '
&
༐ K
༐ P
༐ ༐ &,
༐ ,&&

'?
༐


''༐ Y༐ '&&&
<༐ ^)
༐ ,K:
༐ :,5༐ ,'
༐ &
'
༐༐ K
༐
༐ÖK
&)༐,'Řĉ༐K
&,
༐ ༐?

''
༐
༐༐
&
'༐&5༐
)(༐ 
'

༐ P༐ :
')5 ༐ .)&)༐ ༐
t༐ ༐ ༐ :
༐ ?,&
f༐ Œ')
'S༐


&'[K
༐ ' ༐ '
༐ &'
f༐ ʷ༐ ༐ '
&
5༐ ), ),༐ ?)
༐ e
༐


''
༐K
'
༐༐,& ӆɉ)ӆɉ
'
༐?)
༐
:5
༐5
',?'



''
༐K
༐'
༐༐,&༐ӆɉ)
'
༐?)
༐
:5
༐5
',?'
༐ ,?

t&

 b ¬_ùȮ>༐'༐༐༐
t&
༐Y ?)
༐Y&?'
'Ũ '? f༐])
༐'
&

t&

༐b¬_ùȮ>༐'༐༐༐
t&
༐Y༐?)
༐Y&?'
'Ũ༐'? Y༐
f༐])
༐'
&༐Y༐



༐ 

''



''
༐ &,&,༐ ':
༐ K
༐ '
K༐ '
༐ ?),༐ ,༐ )
'༐
 
P&?'ʗ

༐P&?'ʗ༐
ƞ ’)
༐ K
༐ ༐ K
༐
'
༐༐) )༐
:'
,༐ ?

','
༐ '༐

'༐ &༐),༐.༐ '
༐ ༐ ,K',༐
༐ ,༐
:

༐

'‚
'
>༐ 
' ༐ ༐ ༐ ༐ '
'f༐ ʞt,&༐ 
',? ༐ :
༐ '5༐ Ã༐ K
༐

?
'
<༐ ‰&
༐ 
༐ ?

''
>༐ ).
:
'>༐ ɛ༐༐
t༐ P༐ )
༐ )
'༐
&༐ P༐ ༐ ?K
>༐ ༐༐ '
&

༐ ),?༐ K
༐ %
༐ 
'
༐ ༐ )
༐ K,༐
P༐,?
༐

''
f ĺǖ༐

ྔ೅ೡ೅ྔ೅໵౎೅྿ᛦ

í † ေAေNN$ေ;
ÝNေ ေ NÝေ 
;N
:ေ :$Ýေ :N LေÝ
;
$ေ )
$Ýေ $ေ $ေ 
Å
 )ေ
,7NÝ
 ıေčeေ ™eေ ‫ƒ ם‬ေ ŠC$
N$Ýေ7
6Ýေေ
ေ–
Å Ĩ|,7
ေ÷,Ý )ƒ ć ေ
–N

Nေ C$
N$Ýေ ™$7¨ƒေ + $
Ýေ š e ေ C$
ؕ„ေ
‫׾‬N  ŭေ CÅê$:O™Nေ
V¨ေ+6$)„ေ÷Å e ƒ ေX
:ေ·¨ƒေ901ƒေeေ2̻9íေ ေ2Ĩí1eေAေê
6$ေ $Ýؕ$ Nေ
ေ NÝေ $7¨ေ$ NÅ
ေ:$Ýေ,7NÝ
ေNေ¯$¨,,6Ĩ|
Å N¨„ေ¯ eေ2ေ Fm
7eေ90©¹ ‫ڍ‬ေ
eေ9 0˜í2ေÍ
ေ
Ý
 ေ࣪÷ ,ʼn)ေW
;N
: ҳေNÝေ$ေ
Å
 ေ;
ÝNေ$ေ
u
ÝNေ
ေ
ေ
$N
ေ6$
N$†ေ
eေ Ceေ meေǮ,$
)ƒေ ŠV$ŨေÅ 6N,6Ĩ7$Ý
ေ

Å $ N „ ć ေÅeေ2 ေA,$ေ+Ȝ
;
 N zေA6
æÅ$ေ–
Å $
Ýေ4ÅN
)ƒေ”,
ေ9bí1„ေ90 ƒေeေ21˜eေ
eေ čeေ ™ e ေ‫„  ם‬ေ Š–
Å ေ ,7
Ýေ ေ ÅN NÅ$ေ
;N6
 Ý „ ć ေ 90Iေ ‚W–ေ X$ N$»ေ
+;
 NေW
Ńေ †ေ÷÷‚„ေeေ í2í˜920eေ
2eေ ‫†¯ٮ‬ေ™eေ‫ ם‬ေ$ေeေ ”eေWNÅ
„ေŠ–
Å ေ ,7
ÝေေÅN NÅ$ေ
;N6
 Ý „ ć ေmN$ေ
W
 „ေ4‰$ေW
Ý
$Å ေ‚Ý æ ,
ေေčN ေANေ€
;
6
 ေ+

ေF+࿟
$Å ေAmေ F©9©¹ Ĩ 2©ƒေ”,)ေ90I ŭေčA€+ေ|
Å ʼေW
 eေIĨ2 2„ေȥǞÅLေေÍ
Å NÅ$ေ
4
;NÅ
݄ေVေ9 91ေ FNÅ
ေϠ e¹ ƙေA4͂AေW
 ေXeေA€˅í 91ေ
eေ X^ေ ™eေ |$)ƒေ Š€
ÝNNေ ေ 
N$7NN ) ƒ ć ေ Å ^ေ ‚W–„ေ / F”,
ေ 90Iऽ„ေ eေ
1í 9˜1^ေ
©ʼေ ” e ေȥ ʼ ေCÅX$)ƒေê^ေ ™ ^ ေC
Ý„ေ–eေ Aေ¯

ɩN
%ေ$ေCeေ meေ ™6
݃ေŠ–
Å ေ ,7
Ýေ
ေ
ေ $Ý$ $ NÅေÅ$7
ေÝ)݉
6„ ć ေV
ေ4)Ý ေÍ
Å ʼေ”, ^ „ ေŎƈေ F”$eေ í0I „ေ eေ
9© ˜í11^ေ
Iေ –eေ êeေ W:
„ေ ŠȥေÝ6
ေ$Ý
Å Ýေ ေ ,7
ေ
N$7NN ) „ ć ေŠA;$ěLÝေNေ–
Å NÅÝေ
$ေ–
Å ေ )ÝNÅ݄ćေ¯ ^ေ णࡃˉᛦ ^ေ91©˜ 1ေ F901¹ †ေ
1ေ ” † ေ e ေ‫ٮ‬φ
Ý ƒေ Š|
Å Nĝ,
ÝေေÅษေ

Å NÅေ
ĝ,N6
 ƒ ć ေ  zေ %ေ E &Ա ‚ƪ  à Ա
ƈŕ ij 1I˜0ေ F€
Å  ƒ ေ901 eေ
  !' '

&

D   ü

^)
༐ 
P
༐ P༐ ༐ ༐ ༐ '&'༐ -
'༐ P'༐

&'༐ K
༐ )༐
K

༐
K)
༐:Ł'༐༐
'༐P༐
:
'༐
&
ô༐f
jô༐
:
'༐&
༐)
༐-ֶ
P&' ༐
&)Ŀ
༐P༐ )
༐-༐'༐.
'
༐ Ř
༐:
'༐K༐)
༐ '༐
'. ༐K
༐'༐༐)
༐a7Ũj༐ ’)
༐)
༐''
&༐P༐ ༐༐:K
༐
'༐༐)
༐'

&
༐P༐༐ K༐-༐P༐)
༐-Ř
༐)༐-F
Ș
'༐ )
༐ 
༐ &)&
༐ P'༐ )
༐ &
&
༐ -ô༐ )
'
༐
t༐ ༐ &)༐ &-Ș

 ༐ '
༐ ༐ 
༐ ༐ P'༐༐ '༐ K
j༐ z

ô༐ -༐ &)༐ K
༐ '
༐
.༐K
 ༐ -
༐ P༐ -
ô༐ K༐ ༐ )
༐ -'
༐ ༐ KƗ-'
༐ K
༐
'
༐ ༐ ༐ K
ô༐ ༐ ༐ '
༐ -༐ .
'༐ K
<༐ ^)
༐ '
&
-
༐ P༐
༐ K༐ &
'-&༐ )༐ K

༐ -

༐ ༐ )
༐ ʷ'&'༐ŀ
:
.༐ ༐ ༐

:
'ä '. ༐'
ô༐

&༐P'༐ -&'.:
༐ K
ô༐ '
Ģ
༐ K
ô༐ ༐
K
༐ ༐ K
༐ 
'
༐ ༐ ) )༐ 
-
''
 ō༐ .)༐ &
'༐ ༐ ༐ )
༐
)༐ 

'-
༐ )
༐ 

&༐P༐ ô༐ -
µô༐ '༐&
'-&ô༐ ༐ )
༐ '
P
''
༐

:

༐ -
'༐༐༐ :
༐ 
 ༐ 'K
-ِ༐ z༐ ༐ '
༐ -
'
ô༐ )
༐ &༐
P༐)
༐-
'༐
Pô༐ ༐)༐P༐༐PK'&ô༐.༐K
༐)
༐
&:
༐P&'༐
༐  ༐ ༐  
༐ P༐ )

༐ K
༐ &)&
༐ .༐ P«༐ )
༐ 
'༐
'
Ŀ'
-
j༐
Ŭ
༐༐.༐&
'༐)
༐K&༐ :
༐༐:
༐ P༐ ༐༐
༐ K
༐
:

༐-
'j༐

  
 ð 0ð  ­ð 7  
 ð 0ð ®ð

b j ༐ ^''
&༐༐'༐
aj༐ Ŭ.༐:'༐'
'
༐
™j༐ X)
-&༐
'
༐
Mԓ z-
'-
K༐༐ 
༐
_f༐ ’'ŘK༐༐:'༐)
༐
Ǣʗ༐ ŀ
K
༐'
 )༐
¬j༐ ßK༐P༐ 
 ༐༐-
༐
‹j༐ Ž:K༐༐-༐&Ř༐Ģ
༐
•<༐ ÿ༐༐ '
'
༐
b7j༐ ŀ
K
༐ &༐
b j ༐ Œ' ༐ ༐-&༐
aj༐ ß'༐-
&)&༐'
 )༐༐'
ȸ

&
༐P༐'༐
™<༐ ß'༐)
'-༐)&Ř༐'
F&
༐
Mԓ ŀ
:
༐ ) )༐ 

&'&༐ ༐
P&'༐
_<༐ ë

''༐ ༐ '5
&
༐ P༐ &Ș
&༐'༐

ŗ༐ P༐ )

༐'
'
༐'
Ŀ'
༐ -'
༐

༐&༐K
&
༐ ༐༐

-K
༐ ༐ -Ř
༐༐ ä&:
༐ F
-
༐)༐.༐K
༐'
༐P'༐ )
༐
ढढᛦ
ba༐ ¹ ¹
Ա ¹ Ա + 5
Ա =Ա  Ա  
Ա

-༐
༐༐ ༐)༐
t<༐ǻᛦ ^)
༐. ༐
༐.༐ :
༐)༐-¼ľҺ

༐༐)
༐
'
༐-
༐༐)
༐K༐༐:
༐,༐,:,, 5j༐
Œ'W༐༐.༐K
༐
ľ
'༐༐ľ༐)
༐'
༐༐ S༐K
ľ
༐-༐༐F༐
'
'
༐.༐)
༐K
༐K

'༐
'j༐


üp
ü ü

ü
’)
༐ ༐༐༐༐༐)
༐

༐)༐༐),༐)
༐' ༐ϫƸ)ľ)༐༐

'༐
ľF
༐ .)༐ ༐ K
S༐ ༐ ľB༐ )
༐ ľ༐ ľ'
༐ 'ľ'
༐
.)ľ)༐༐ľ),'ľ
'ľ༐༐-
,j༐ ^)
༐)'ī' ŁS༐ '
'
༐ -,US༐ .)༐
.)ľ)༐ ༐ľ'
༐ 'ľ'
༐
t༐ ༐ W༐ '
༐ ༐ K[-ľ'ľľ༐ -
ŧ
W༐ ༐ ༐ )༐ '

ľ༐ ༐ ༐ :
'༐ -'༐ ༐ ༐ Ŀ<༐ ’)
༐ -
༐
༐༐ľ
Æ༐༐

ľ༐.'༐ľ'‘༐
t༐:
'༐༐'༐''.༐
'
༐ ༐ F௅-
''
j༐ ^)
༐ ',༐ ª'

༐ ༐:ľ༐'
:
༐ )
༐ '-r
༐ ༐ ,'
༐ ľ'༐ -S༐ ༐ )༐ ༐ ƍľ༐ )Ł༐ Ŀ༐ 
༐ ༐
ƞ'‘
༐ < ĉ ༐ÿ༐ ༐)༐)
'--ľ༐FKŁ༐ ༐ 5U༐ ༐ľ'ŧ
‘
༐.)5༐༐༐)
༐ ༐)
༐'5'༐
-
','
༐'༐༐ķľ౻
༐ ༐-
j༐
^)༐ )(-
༐ ༐ ľ'F‘༐ ༐ľ,
༐  Ø2
ð ༐ '

F༐
༐
:
'}'

༐ )'
༐ ༐ )
༐ ,༐ .'B
'f༐ ^)
༐ '
༐ ༐ 
-5''
༐
.)༐ .)ľ)༐ 
:F'ƍľ,༐ ༐ BŁ༐ ༐ B
༐ ľ
༐ ༐ 
ľƍľ༐ '༐ é
'
༐
༐ ľ-༐ ༐ .༐ K5༐ 

'-
༐ K༐ )Ł༐ :
༐ ༐ :ľ༐ ༐
ľ'‘༐ :
ľ<༐ Œ'༐ ľ'‘༐ F༐ ,B5༐ ľ
W༐ )
༐ :ľ༐
-༐KŁ༐.༐
 )༐ ÐK༐ y7Ĩ༐
© ˈ༐༐
'-༐ '
༐༐.)
:
'༐
)'['
༐ ľ'༐ -༐
t<༐ z༐ ',ľľ
Æ༐ )༐ 
:'ƍľ༐'
༐ ༐

༐ĿľB༐.)
༐༐ ༐-
༐༐ľ
S༐ 'Æ༐ '༐),༐-
'S༐.)
༐ ,༐
.'
༐ ༐)
F
༐ ༐)
༐K.༐'ľ)༐K༐)
༐ ༐.'B
'<༐
z༐༐
༐༐)5༐ľ)'ľ
'ľ༐༐ ༐)༐༐
༐༐
༐༐ )'༐

ƍ
༐ -
 ༐ ༐ .)
'
༐ Ŀ༐ ༐ ༐ )
༐
tƷ༐ ༐ 5ĿK'- j༐
ÿ༐ 
༐ F༐ [B
༐ ľ)'ľ
'༐ 'S༐ ༐ )
 S༐ K༐ :'5༐ ༐ ༐
ľ༐ 
ľ'Ł
༐ ༐)
༐ :
༐ ༐ )
༐:ľf༐
Ǻᛦ ࢚
/E
ေ,/
ေေ+ɲKေK//
/ေQ
/ေ=ေ
/7ä
ေ=ေ#ä)ေ=
ေ#
ေ,7
/ေ
7)ေ: # ေ= L)ေ඲ေ /E
#ǟ
%ေ ေ=ေ#ä
ေ,Eေ /
ေ /##
E=ˋ/ေ : # ေE
;ေ ေ

Kေ=ေ=
ေ+Kေ=
Qေd)
u ßေÍ /ေLေ/ေEE
ေ=ေ/
ေ®&ေĂǍ

=ေ
K//ေ #E/Ë/ေ ေ 
/ေ = ,/ေ =ေ #;
)ေ =
ေ #E/=ေ ေ ေ Kä//%ေ ä= ,K ေ
)ေ,= /ေ,/
ေ=
ေ
ေdð)
uoေ:= ေ=
ေE#=ေ= ]=ေ=ေL
/ေ=ေ# L#)ေ

/ˋ/"=ေK//ေŗI2& ‫܆‬ေ " /ေE#=#
ေ/ေ=ေə ,/=ĂĞ
ေ Í
ေ )ေ#
ɂ#,=ေ 
/K=ˋေေ
+KေK//
/ေ /ေ=
ေ,̻,7
ေ#
ေ: # ေ ]/ေ7

ေေ,/
ေေ/
ေ=
ေ Í °ေ
ေ d7=)ေ#
ßေ འေ 7

ေ /#=,
‫ݸ‬ေ)
/ေ K%ေ ေ )ေ=
ေ ,O,7
ေ
#
ေ:ေ7
ေ,/
ေေ= /ေ 7Q%ေ
;
ေ:
ေ


#5/ေ ]Lေ ĝ,
ေ: # ေ,/
ေ "
ေ ေ

/K=/ေေ=
ေKေ=
u= ֎ေmေ=
ေ#ę=ေေdäေKä//ေ:Q
/ % ć ေေ"ေ;ေ
#,/%ေ 7= ေ#
/ေ
ေ
E,#
ေေ= LေAEE
u%ေ ೧7ä
ေA ˅  ေÍχä
ေAေKศ;
/ေ=
ေ
#
ေ,7
/ေေ›ˋ7LေK//
/ေ: # ေ
ေေ#,/
ေ/ေ:
)ေ,/
 ေ
̯ေÍ
ေ ,=ေęေ;/#/=)ေ/ေ 
ေ dE/
oေ : # ေ /ေ =
ေ 
/ေ ̹ϨƒҫŒԱ Ŀေ=
ေ
# OKO/ေ/)/
ေေ,=/ေ=ေ/ေ= ,/ေ
/,
ေေ)
/ĕ/
#ĕ#ɪေlä)ěLäေ ]/ေေ;/#/=)ေေ
9&ေE/
/ေ=ေƅå+èေK
ေ/),Eေ¥&ေE/
/ေ=ေ&å+%ေ E=# ေ9&ϽေE/
/ေ=ေå+Üေ:"
ေ
&&9ေE/
/ေ=ေ&Ħ+ÜေေNေ&í1¥ေ ®Ա 9&ր ϽေE/
/ေ=ေ91Ħ+Žေ
 \(ÏÏԱ

M<C 
Ǧ



0
Ǧ Ǧ  Ǧ Ǧ  Ǧ (0Cš
Ǧ Ǧ Ǧ  

Ǧ

Ť2~Ǧ ž 
Ǧ NBïƶ Ǧ 

0É¢,Ǧ STǦ Ǧ 0
v<š<CŖŎŽxǦ X <CǦ Ǧ ;
Ǧ
ƖSƷÊòöûǦ XÁ“Ƙ
0ÐNǦ ž Ǧ  Ǧ TSǦ hǦ _f2Ǧ 
Ǧ
i0Ǧ NǦ Ǧ
Ǧ
Ǧ 
Ǧ  _Ǧ İz 0N LǦ ŻǦ lj z
ČǦ y 0,Ǧ T2Ǧ Ǧ 7
:Ǧ
mĈddǦ ÐǦ Ƹŏǝļ¤Ǧ 0Ǧ NX0Ǧ Bß0ǒ Ǧ Ǧ
LǦ Ǧ bį,0,Ǧ h
Ǧ § dĒĞǦ Ǧ § d¨ ĥ Ǧ
0NN5Ǧ 2ƭ
8ƧǦ 0
Ǧ zǦ Ǧ 
—Ǧ 0

 NǦ Ǧ ;Ǧ  
 8Ǧ É¢(ƼǦ
Xg“¸_ǦǦ SǦƦ ,ĬǦ
ǦĜCđ Ǧ
tǦǦ  pǦu 
Ǧ ttǦ

Ǧ
Ǧ
Ǧ 0
0 Ǧ <šǦ ;Ǧ à

Ǧ
5~Ǧ pp5Ǧ 5Ǧ

h

BðǦ <C(5Ǧ  Ǧ

5Ǧ  


Ǧ
Bñ5Ǧ 
GǦ D Ǧ 

Ǧ

,Ǧ TǦ 
—8Ǧ À”

ŶǦ Ǧ  Ǧ  
 Ǧ 
Ǧ Ǧ  Ǧ i ĸǦ ÎǦ
gž
LǦ

şǦ Ǘ ƨǦ
, Ǧ Ǧ ’Ɨ
Ǧ _Ǧ Ǧ
i0Ǧ
Ž<Cx5Ǧ h:Ǧ 
Ǧ
Ǧ ƺ›
ĽÊC5Ǧ 02Ǧ LǦ Bá  _TǦ  »Ǧ  ,xǦ Ǧ 
<ËǦ Ǧ 
dž
Ǧ
Ǧ 
 5Ǧ 
Ǧ
80L5Ǧ
Ǧż”
_Ǧ
,čǦ
H Ǧ 
 Ǧ 
Ǧ  —NǦ bǦ Ǧ  Ǧ  Ǧ Ǧ 0Ǧ Ǧ
ƚ5Ǧ 5Ǧ  Ǧ


ǃǦ L(Ǧ Ǧ Ǧ  Ǧ  Ǧ Ǧ Ǧ  Ǧ
Ǧ hǦ

 Ǧ Ǧ

Ǧ Ǧ



Ǧ Ʃ Ǧ T

Ǧ Ǧ
Ǧ
t005Ǧ Ǧ Ŋ›
Ǧ bǦ




GǦ ?:Ǧ ƾ ‹  ǯ 8 
Ǧ  Ǧ Áƙ<Ë

Ǧ Ǧ  Ǧ
 Ǧ
Ǧ 
Ǧ  Ǧ

: Ǧ
ǦǦ Ǧ( <CbǦǦ Ǧ 2 pǦ
H 7Ǧ ſƀ
Ǧ b0:ǦǦ x,ǦǦ  LǦ Ǧ
 5Ǧ ž <CǦǦ;
 Ǧ
:Ǧ Ǧ À”
5Ǧ Ǧ  ÄƹǦ Ǧ ?:Ǧ  . 5Ǧ *Ǧ y85Ǧ Ō 5Ǧ <CǦ ƪŽ
P

ţƈǞ ŧ੢
!)

'\

\

\ \ \ &\ U

$\ '\ * 0\
'ǞŤ%?%?l(Ǟ  ,CK\ a ʇ੢ĸŧ࣡ࣾ ʸ੢
Ã/ç
Ƕ ገ ‫ أ‬ገ ‫ أ‬ᛦ 6Ï
›/è
Ƕ › / 
3Ƕ
` !/Ƕ ÚǶ ƢǶ 
Ƕ 3 X#
/Ƕ  *
;Ƕ L#Q
Ƕ
Á ǑǶ bv5
üП YŁ
5%PП Fü%x
П £П ?ü%Pü ïП
 ӗ Ә ດકອ
ਖ਼ອ
৽߅ᛦ ਗ਼ອ

ৌ߅ᛦ
ອ ‫ے‬ອ
ɶ
‫ۓ‬ອ
ேைᛦ Īອ
<ອ Īອ
‫۔‬ອ
؋¸ອ ॵ߅ᛦ
‫ڒڑڐ‬ອ
Łॉອ
ঙອ
ᙾᛦ
ņᛦ
°ອ ࢟߅ᛦ
ोອ
Ƅອ
ອ
،¸ອ
ŀອ

ࢠ ߅߅ᛦ ॵ ߅߅ᛦ ৌ߅߅ᛦ ৽߅߅ᛦ

-8 F L 8 O3UYP8a Јͬ ξԓ
Ä/4@IǶ Ĵ ī· ĈĵĶ† Ƕ 6ƎW4IIǶ GǶQ 
3
Ƕ /XN 3ĔăǶ #Ƕ Ƕ ǶI`LJ X/Ƕ @ Q
Ƕ L#ǘ
3Ƕ“W ĕǶ
ũ W )5Ƕ ǶĬƉ33/ Ǖ Ū ¨ Ƕ ^Ļ3ÁP[щĊԱɈ ӕԋƲšӈţԱ PžӉ ƀ&Ա
3Ħš [Ċ-…Ա

৽߅ᛦ
E
%E
E
E
E
ৌ߅ᛦ ‫۽‬ອ
੘ອ
۷ອ
ෘᛦ
ಋᛦ ާᛦ
ອ
ɷ ިᛦ
௅ெᛦ ॵ߅ᛦ ੗ອ
ՐŅᛦ ۸ອ
۹ອ
‫¸؍‬ອ ĩອ
ࡊ ‫ۻ‬ອ
‫ۺ‬ອ
<ອ
°ອ ĩອ
࢟߅ᛦ ‫ۼ‬ອ డᛦ
Ցᗍᛦ E
E
°ອ
E
Łॊອ E
Ƅອ E

E
ອ E
‫۝‬ອ E
ŀອ ߅ᛦ E

E
E
E
E
E
E
E
͙ ࢟ ߅ᛦ
E

࢟ ߅ ߅ᛦ ॵ߅߅ᛦ ৌ߅߅ᛦ ৽߅߅ᛦ

-8 F L 8 P3VYP8a ʆ ‫ މ گ‬ʷ੢
Ä/43
Ƕ < C =ǶmO Ƕ 6Ɩ4ťŸªǴIǶ GǶǖǶI3
Ƕ×dǗ3RǶ #Ƕ  XXǶ Ƕ
` X2Ƕ3 ƧŸ Ƕ L#N3ǶMW [Ƕ
ÅƵǦ:Ƕ ǶǵŤ˜33/Q   ƶ Ƕ ¼Ǿˉ3ʼnP[ ĊԱ( Ƅ ²š&ţԱ )ӊšъԱ
3&c [Ċ…Ա

डਚᛦ
 \


Ա ‫_ڭ‬༐

'?ôӳ༐ Ö.'Ř ༐ .?)༐ ༐ -*-༐ ȡ༐ ô༐ǻᛦ '**༐ ,-*༐ ༐ )5༐ ,༐
༐༐ -K*'༐༐e*-*''*༐K5.**༐ ॱȉǢ7ùX༐༐Ǣa7ùXô༐)*༐-*༐༐'*?r
-*༐ :' ༐ '-༐ ༐ *.༐ -*༐ ༐ )*༐ ) )*༐ *-*','*༐ ༐ *:*',༐
.5*؁༐༐)*༐.*༐*-*'?'5<༐^)5༐-*༐'**༐༐Ǣa7ù༐.,༐Ŀ*)*༐
༐.,e*'ô༐ )*༐'*-*'༐,'})* <༐^)*༐5T༐'*'*༐༐)*༐,-*༐
.5'5༐ )*༐ **'-*ô༐  ༐ ༐ )* ༐ ',*༐ ༐ їIJùX༐ *'༐ -*༐ '-༐ '-༐
5-*''*༐ e༐ ‡‡˲ȏùXô༐ ?)5༐ ‫׀‬-*༐ K* ༐ )*ſ༐ ,?༐ ?)༐ *-5',?'*༐ '༐ ༐
-*༐ ķ*༐ ſ ༐ ༐ ,?,༐ ༐ ?,༐ ſ* )ô༐ ),༐ ô༐ ,༐ * )༐ .))༐
.༐?༐), *༐.)༐')5'༐) ༐༐eÓ*༐-5༐*-*','5 ̔༐¼,ſ༐)5༐
-5༐ƌȗ,༐*༐,?༐,'T-*༐ѕIJùX༐*'༐-?*༐?༐'-༐*-*',?'*<༐Ž༐
**?༐'-༐)5༐ '*?༐༐).༐༐Œ '* ˹༐<a<༐X':*༐Ž༐).༐)*༐'5༐
K*༐ ༐ )*༐ ,-*༐ '*:ſ༐ '*,?*༐ ༐ Ǣa7ùX༐ '༐ y_༐ -?5༐ ༐
Ŀ*)*༐ ༐ ƌԱ?*'<༐ z༐ )༐ ,*ô༐ ,༐ 5-*''5༐ KQ*༐ ,K༐ a_7ùô༐ ༐
',༐ .,༐ *'-*༐ ༐ ?)5༐ '-J༐ 5T:ô༐ ?)*༐ '?*༐ ༐
)༐?'?༐'5, ༐ƌ֭?)༐'*ΞĹ5༐༐?*-*''5༐༐eÓ༐༐*-*'‚
'*༐ ,K:5༐ ,K๏༐ їIJƀ7ù༐ )*༐ ༐ )༐ ,༐ '*༐ * :5༐ *T༐
*ķ5 ̔༐ ༐ )*༐ ) ༐ *-*',?'*༐ ༐ ‡‡їIJù༐ )*༐ 'e༐ *༐
'༐ -*༐ -*༐ ?J༐ )5༐ ༐ -*༐ ',)*༐ *ĿK'-༐ ?༐ e)༐ 5-̨
*''*<༐ X'Q*༐ ¶༐ ).༐ )*༐ '*༐ K*༐ .?)༐ ௣ອ -*༐ '*:༐
'ʽ3?5༐ ༐ їIJ௺܎èùX༐'༐ NQ੢ ༐ ༐ *༐ ༐ ,'f༐ z༐)༐ ݇࢒Ĺ*ô༐ ) ༐'༐ y༐
)'༐ ,?༐ ‡‡ѕȏùXĕ༐ ,5'༐ )*༐ '-ſ༐ *T,:༐ -5'*-*ô༐ '*༐ ,༐
-'Ř*༐ ½'5ĝ5༐ ༐ * )༐ ༐ ?,༐ ?5-*',?'*<༐ ^)*༐  ༐ ':5༐
K*ô༐. ༐*,)༐༐?)*༐*T:༐':5༐Ž༐,༐ OąƱ ,'5༐,ſ༐).ô༐
,༐ ,'*༐ *'༐ :*'༐ *,'༐ ?)*༐ -*<༐ ¯༐ -ſ*5༐ *ĿK'-༐ K**༐
K?,*༐ ༐ ‡‡ѕӷùX༐ ༐K)༐ *ô༐ ༐ K༐eÓ*༐  ༐ ':*༐ .ſ༐ ):5༐
K**༐*,ſ< ĉӶ༐
Ň^कį༐ )༐ '**?༐ ,5-5༐ Φ༐*''*e?༐ ༐ '*?༐ ༐ )*༐ KQ*༐
*༐ ༐ ?)*༐ Ǔ༐ ༐ )ɚ༐ 5༐ ༐ ࠀ¼:5Ũ༐ ?5-*''5<༐ ^ *༐ 5༐ ༐
ůʅ?ƌɉ5༐ *-*''*༐ ༐ '5༐༐'*'༐༐ Q5༐ -,)*-,e3ſ༐ 5T'*༐
?༐ ?)5༐ ,‫ۀ‬༐ ),༐ ?)Ï༐  })*-,ſ༐ ,?*༐ '༐ ༐ ༐ ,༐ ,༐ ,?༐ ௢ອ
Q5༐ *-*'3'*༐ ༐ e)5༐ È'ǒ!༐ '* ½༐  ,Ɲ5༐ .)༐ 6&-"9 ,?༐ Ǫ༐
eÏ-*'?Ű*༐ e༐ 5dK'-༐ ༐ '*)5 <༐ ʒ)5༐ )༐ )[)ǵ-,༐
5dK'-༐ ༐ ,?,*ô༐ eÓ*༐ ƺ Q5༐ e5-*',?'5༐ ༐ 5Ŀ༐ ?༐ eÓ5༐ ,༐
?*-*'?'5<༐ɑ5'*༐*ĿK'-༐༐'5)5Ʊ༐ )ધ‫ࠁڪ‬5ƌШ*'ô༐ ?)*༐)Èˎ 5-,༐
Ȥe,?*༐ ༐ ?)5༐ ݇ࣄĹ༐ ''*༐ ༐ )*༐ *dſK'-༐ ,e*༐ !༐ -*༐ Çē5'*༐
ư) )5'༐'༐.*'Ě༐ *-5',?Ű* Ɗ༐ి༐í>༐༐)*༐ ༐.*'5༐)*,?*༐:*'༐dŘ༐
ອ “Ƕ 
Ƕ ¦ÒƤǶ  !! øǶ d!Ƕ 4!Ƕ Ƕ Ƕ R! ňǶ < †Ƕ Ƕ 
*ƏƻǶ RǶ 
Ƕ  !!®Ƕ Ƕ a·¬Ƕ

N  Ƕހ‘Ė
Ƕ Ƕ N Ƕ  YYǶ Ƕ Ƕ Ƕ 
GHƁr
Ƕ *
…Ƕ *Ƕ 
Ƕ Y
!rǶ Ƕ
ƽƷ!Ƕ d n  Ƕ ƒd*¢Ƕ  !!
YǶ  d*Ƕ Ƕ…Ƕ ¦ÁǶ  Y! : ù Ƕ*Ƕ
Ƕ Ƕ:Ƕ  XǶ
Ƕ Ƕ
Gʼnǡ2r
Ƕ  
…Ƕ Ƕ !
W Ƹ Ƕ!
!r
Ƕ 
Ƕ DSÛ 

S!W  Ƕ Ž!!Ƕ!Ƕ Ƕ R Ƕ
nH‘ųƹƺǶ  !! kǶ mRǶ *Ƕ Ƕ 3
Ƕ … 
:Ƕ *
Ƕ
RǶ mǶ mNj! W  ¢
Ƕ * ! : Ƕ
ƣç
ǶV! [Ƕ    Ƕ
r
*ǶǶ m3 
Ƕ *  Ƕ
Eǯ ΂ɔȯ ¿ЉȯŹ
Ա ( Ա Ļ— 5
Ա ™;¿Ա ; Ա ż ŅĹ
Ա

Ʊ 2Ʊ ŧ

Ʊ&


ƱƱƱ7ƱƱ Ʊ 7ƱƱ7E Ʊ<Ʊ2Ʊ


7(7MƱ 3
ƱƱ-Ʊ 
Ʊ õJ3
öƱ &7 

Q ¦ }Ʊ
-
2
Ʊ 33
2Ʊ(
Ʊ 7Ʊ  Ʊ Ʊ7Ʊ A

<&Ʊ Ʊ 7Ʊ7ưƠƱ
Ʊ (22Ʊ (Ʊ 2Ʊ ñà„Ʊ 3(371Ʊ _Ʊ 7Ʊ 
&<
Ʊ 2MƱ 2

Ʊ 


3
2Ʊ
‡Ʋ•j༐
_ Ʊ /ē<Ʊ <<2iMƱ JiƱ 
Ʊ ´Ʊ ÷Ŗ35”ÝƱ Ʊ ŌŊæ
ƃŏƱ ¡Ʊ Ʊ
Ö<MƱ “52MƱ Ʊ 4
MƄ£ƱƱ7ƱƱ øŗ7ğĨƱ ÔáºƱ <ƱĿiƫ
3
£ĩĆƱ†ÅƱ<Ʊ

3(Ʊ 
32ƱƱÚ(3(i7ć†ÇƱ&7Ʊ 337Ʊ<(ŒƱ


Ʊ &•Ʊ ò(22Ʊ 2ƅ3Ʊ 2Ʊ Ʊ  Ʊ 7ŷƱ Ʊ ŘƟ3Ʊ
<3Ʊ i3Ʊ ƜƱ
A
(
Ʊ 2
Ʊ Ʊ Ʊ /
2MƱ 7<3Ʊ /
Ʊ ´
Ʊ ´Ʊ AƱ
232NƱ 7Ʊ Ʊ 2Ʊ&Ʊ Ʊ /
Ʊ ¡Ʊ Ʊ <
ŠijƆƱ 2¡ZƱ
-
Ʊ 7(AƱ Ʊ 
Ʊ 7<ţ(Ʊ Ʊ &Ʊ 
i
3Ʊ Ʊ Ʊ 
æƱ
Ʊ (Ʊ 7Ʊ 373Ʊ HFŐFHƱƱ 3<Ʊ Ů
Ʊ (33(Ʊ < ZƱ ×2Ʊ
iƱ 
Ʊ 2 
2Ʊ 37Ʊ Ʊ ÌŒ3<Ʊ (2İƱ 

Ʊ  Ʊ Ʊ 2
3<Ƈ¡Ʊ (Ķ2 ĴĖºƱ
3Ʊ óùƱ 3 

Ʊ
/<Ʊ 
Ʊ /Ĕ<Ʊ ”Ý3Ʊ (7ƕ 1Ʊ -
7Ʊ ĵƱ ¶ອ
7²A
Ʊ i3

Ʊ Ʊ Š3
Ʊ 
ZƱ Ł3Ʊ ( Ʊ <iƱ<Ʊ 
²
Ʊ  
Ʊ Ʊ

<
2Ʊ 
ŠƱ (Ʊ Ʊ 
²&
Ʊ 2&(37MƱ Ʊ Ʊ Ʊ Ʊ Š3
Ʊ 

&ƖƱ
Š
Ʊ 
A3(Ʊ (Ʊ 
(Ʊ 
2Ʊ Ʊ 77Ʊ 
22 ÐƱ (Ʊ 
37Ʊ &
Ʊ

<Ʊ 2Ʊ Ʊ 32(7 ĈƱ ×2Ʊ 2Ʊ
2ķ
ŒƱ A
Ʊ 3
Ʊ Ʊ Ʊ /
Ʊ 22Ʊ
Ʊ3Ʊ Ú(3<27Ʊ (
Ʊ  Ʊ 327(AƱ Ʊ 
Ʊ 37 3Ʊ 3 2Ƭ
1 „ Ç Ʊ
_7Ʊ  ‹ÑÉǯ 23
<3<Ʊ Œ2A(A
2Ʊ   Ʊ Ʊ 
ƌƋľƱ &•ŤƱ 7(Ʊ
Ɨ3
Ʊ Ʊ <Ʊ (22Ʊ (2Ʊ2&i
Ʊ  Ʊ 7Ʊ 
á(Ʊ 77<Ʊ 2A3A
Ʊ
Ʊ <Ʊ 3ĸƐgƱ _
Ʊ ʱԓ Š <Ʊ 2ś&72Ʊ &7Ʊ 23
Ʊ Ʊ <Ʊ Š<ĕiÜ<Ʊ
(Ì2MƱ (A
Ʊ ú
Ʊ (Ʊ O2Ü
QíÅƱ

Ã/Ǡnj
Ƕ % nÿ¨ ǖ͕ DoI
/Ƕ 

¡I*ǝã2DŽƾǶ d*Ƕ N Ƕ Ƃƥ/—*Y2 ‘ƼƘƦǶ GǶ ǶÏ2ÓƐ

*Ƕ ž
Ǣ

*Ƕ
N   Ƕ ¡ã'4@–Ƕ #Ƕ ǙǚǶ )¡ Ƕ ĉ #Ƕ *Ƕ Ƕ  YYǶ a@/N † ė ǶL#Q
Ƕ Í 2 Ę
* ; ķ Ń Ƕ ‡ćþǠ Ǖ¶•š ñǯ
ĢĮ ƺįƻƍ™cƎ ǯ •Lj~ ;ǯ”€•'C2Nǯ
 \Ź(ÏÏԱ Ïǯ

ħ Ĩ ǯ Ή ǵȶΓ
:=̈ž̃ ͕ % Z͔ ! ͕ ʒ‚ ē‹ƒ{͕  
ʁʊ¥vv=¥͕ =͕ G͕ Q= =¥͕ µ͕ ठᛦ‚ ͕ bv ƄĚƌ ͕ ,Å͕ ,Qͅ
=5=͕ ɰ̂ēɛ ͕ ɅµĢ ͕ ůƌ” ¥͕ ¥Q͕ɇ={,ˇƯ  ʨ Ʈʩ͕ ˤǾˊ3ʆʄP[ƥÂԱ ˜¢ÅV̿}П ʿ¢αβϿLjǸП Ï3ɱšƥʔˆñԱ

ÖQ$ǦÏ 7ǦǦ

Ǧ ,Ǧ

¶Ǧ Ǧ Ǧ
1Ǧ Ǧ1(Ǧ Ǧ
1Ǧ ĭ åœǦ Ǧ

,þǦ7Ǧ Ǧ 
ǦǦ $Ǧ g{Ǧ 
Ǧ Ǧ  Ǧ


 
Ǧ S,Ǧ Ǧ  Ǧ 
¥1 Ǧ  ŲP1ǘǦ -1
3 GǦ y SǦ
ĮǦ ¼Ǧ  Ǧ BBæ3Ǧ 1$
 8Ǧ Ǧ BèǞ
Ǧ 1Ǧ Ǧ  SǦ
ŧǤǕ Ǧ Ǧ
7

Ǧ Ï1nǦnǦ
Ǧ Ǧ’Ɯ
ǦǦǦ8


Ǧ’ƌ )Ǧ;ǦǦ
Ǧ
S Ǧ 
Ǧ jǦ Ǧ  Ǧ 1Ǧ  
)Ǧ Ǧ ;Ǧ8


ÈǦ

Ó
× Ǧ ǦǦ Ǧ  ǦµµçǦ 
GǦ
ØH Ǧ 1-
{Ǧ Ǧ  Ǧ {Ǧ 1Ǧ Ǧ  
Ǧ 2 ǦǦ ǦǦ 1Ǧ
mŕ77ǦGǦ „dOǦ,Ǧ;Ǧ
ą ǦǦ
3ħǦ Ǧ  Ǧ  
P(2Ǧ 
ŝŚ÷
3Ǧ
 Ǧ
¼
Ǧ œǦ {

$Ǧ 
Ǧ Ǧ (Ǧ Ǧ 7Ǧ 1Ǧ  Ǧ 1Ǧ È
 \Ǧ
Ţ,Ǧ 
Ǧ Ǧ 1
Ǧ 1Ǧ  Ǧ `$Ǧ 
Ǧ .
3nǦ ;Ǧ 


Ǧ 
Ǧ ,Ǧ

ƍ Ǧ T‘ŨǦ Ǧ 
Ǧ 1GǦ â w
Ǧ  
Ǧ FǦ Ǧ 
Ǧ Ǧ
|Ǧ
Ǧ  Ǧ  ¶)Ǧ 1Ǧ ሹၦโᛦ Ǧ 
Ǧ (
 Ǧ 
 *Ǧ H 
Ǧ 
Ǧ ÄǦ Ùã Ǧ
- Æ Ǧ '
Ǧ 2
Ǧ Ǧ /3F)Ǧ Ǧ l 
Ǧ ƈ/Ǧ 
Ǧ
3Ǧ
Ǧ ]

Ǧ
|Ǧ Ǧ
1FǦ [
¹Ǧ Ǧ 

F*Ǧ ł
Ǧ Ǧ Ǧ
8Ǧ 
Ǧ
- Ǧ 28¯FǦ 
8 Ǧ $ÿǦ
 TǦ  Ǧ [Ǧ Ǧ gÇǦ 8Ǧ 1Ǧ
3ǦǦ ]
F8 * ä Ǧɜ ŧ ອ
D Ǧ $P
 Ǧ
1Pl1
3Ǧ  Ǧ |
Ǧ   Ǧ 

Ǧ
Ǧ ǖŗǔ

Ǧ Ǧ Ǧ Ǧ 
 Ǧ 

 Ǧ Ǧ ( Ǧ  Ǧ  

 Ǧ


ǦǦ 8oǦ Ǧ[
 Ǧb1
Ǧ 'Ǧ

 Ǧ
FǦg½
Ô
|Ǧ
ůƎ U ¨ ­ Ǧ RǦ 
¸
Ǧ 
nǦ Ǧ  FǦ

)Ǧ Ǧ  Ǧ ƛǦ 
¦
˹‹༐ Å
ϔϕÏԱ
Ա Ļ— 5ÏԱ ™;kԱ Ļ; Ա ÏԱ

Ǧ Ǧ D Ɲ­ţǦ -Ǧ Ǧ ƞ^Ǧ Ǧ (i·Ǧ LǦ Ǧ &;ºǦ 

»Ǧ  ŷ“Ơ Ǧ


Ǧ Ǧ 2 Ǧ Ǧ $$9Ǧ Ǧ 9ƏǙ
3Ǧ Ǧ ũL Ǧ
ǍǦ Ǧ +
Õ

Ǧ -Ǧ ƒ ğPFǦ
 
Ǧ Ū'&Æ*Ǧ QǦ Ű&
Ǧ Ǧ ųǦ ū
 Ǧ ŬƿÕ
ǦǦ
 +Ǧ Ǧ Ÿ'&
 * ~ Ǧ


üM8  ü üü

‹ Ǧ Ǧ Ǧ Ǧ  9Ǧ

 JǦ &Ǧ Ǧ

 Ǧ (Ǧ ഒಳˈອ

JǦ mǦ 
O # Ǧ-
Ǧ
ǐ› ,Ǧ Ǧ Ǧ


Ǧ  Ǧ Ǧ (Ą \Ǧ
&ŴǦEǦ Ƒ‘,Ǧ
ƽǦ
 ĩǦ ‡E=%Ǧ #Ǧ œ%ĺÅǦ #ǦQ–WīǦ#Ǧ VW#Ǧ ŐR#Ǧ ªŠ%=#Ǧ ǟ%=#Ǧ
Ǧ 
Ǧ Ǧ Ǧ 9Ǧ 6
 *Ǧ ǥ Ǧ Ǧ 


Ǧ Ǧ Ǧ ǀǏ
Ǧ
(EǦ Ǧ 
Ǧ

$ 9Ǧ 
Ǧ ƻsǦ +Ǧ ^( 
#Ǧ '
Ǧ  Ǧ
Ǧ Ǧ Ǧ
Ǧ Ǧ ,ǦǦ
9Ǧ &Ǧ 
' •Ǧ -Ǧ Ǧ 
Ǧ Ǧ iǦ :Ǧ
Ǧ 
Ǧ  ǦǦ &

Ǧ -Ǧ 
Ǧ 
ǦǦ 
Ǧ Ǧ ÍĉǦ
‹ Ưs•

Ǧ 
Ǧ (EǦ -
#Ǧ Ǧ  '#Ǧ 
Ǧ  Ǧ +š
Ƌs \Ǧ
+ǦǦ] Ǧ

J ˆǦ‡>%Ǧ#ǦªĢ%υǦ# ĕ¡%=r #Ǧ ‹+>%Ǧ # ǦM%>…Ǧ DǦ -Ǧ
ć֦

(Ǧ- Ǧ Ǧ
Ǧ8ǦlǦ Ǧ-ǦǦ
ǦǦǦǦ Ǧ

Ǧ (+ ĪǦ D€ƉǦ#Ǧ Ř%€ør #Ǧ M–>¬ #Ǧ IJ©ď= … #Ǧ ´>¬ #Ǧ Ǧ D=% *Ǧ DǦ ƅ
$Ǧ
&
Ǧ (LJǦ NJǦ ÎǦ 6 Ǧ 
Ǧ  #Ǧ Ǧ Ǧ Ǧ
+Ǧ
Ǧ  Ǧ
Āé
Ǧ$ Ǧ  6$, \Ǧ Ž,Ǧ D+=%Ǧ 9Ǧś%>rǦ:Ǧ
Ǧ  Ǧ  ,Ǧ ācǦ
JǦ -Ǧ ľWǦ:Ǧ Ė¡W#Ǧ Ǧ Q–ĻÅǦ 
ǦǦ  Ǧ  ưs°Ǧ -Ǧ VW*Ǧǻᛦ
ŸR
ǦǦ Ǧ(#Ǧ  Ǧ&;Ǧ Ǧ Ǧ 3J°ǦǦJ3 ^Ǧ
#Ǧ
3Ǧ Ǧ 
Ǧ EǦ &Ǧ 2 Ǧ &-
3*Ǧ ´+Ǧ ƫ %>Ǧ Ǧ #Ǧ '
Ǧ (Ǧ

Ǧ 
Ǧ Ǧ Ǧ &
º
 Ǧ

Ǧ Ǧ ˜êǦ (,Ǧ 
Ǧ ŵ
Ǧ
Ú 
 *Ǧ D$Ǧ ]F,,Ǧ Ǧ Ǧ Ǧ &Ǧ 
2Ǧ Ǧ 2Ǧ w>ĤP
& 
+
‘ǦǦz
3#ǦǦ-9ǦLǦ?
Ǧ 
.QǦǦǦǦ
^Ǧ3-˜ŅǦ
ആອ 9-
3Ǧ¾
6* Û Ǧ

±&
 &Ǧ 

Ǧ Ǧ E&Ǧ -
Ǧ -9Ǧ 
 Ǧ &-:3Ǧ



ǦǦ
 Ǧ-
3Ǧ¾
 ijǦ Ǧ
Ǧ

Ǧ Ǧ ŔR#Ǧ W#Ǧ
VR#Ǧ őR#ǦǦ&
Ǧ(EGǦ
±:^Ǧ 
Ǧ EǦ 2 Ǧ Ǧ 
EǦ 9Ǧ Ǧ 
 Ǧ ńƮ
Ǧ
ǚ
9+9&,ǦǦ 
Ǧ

 Ǧ 
Ǧ
 Ǧ

9Ǧ -&;Ǧ Ǧ 3Ǧ - ťǦ Ǧ
8'Ǧ 
:ǎ \Ǧ 6+Ǧ 
Ǧ &Ǧ 
Ǧ Ǧ  Ǧ 
Ǧ ?%€ ƊǦ #Ǧ
ĝ©9%= ĴĵǦ#Ǧ ‡=ġǦ#Ǧ ?#Ǧ Ǧ
$*Ǧ ³ǁƟ•Ǧ &ǦǦ 
ǦǦ Ǧ

Ǧ

 :Ǧ J·$$G ē Ĩ Ǧ ?
Ǧ &Ǧ 



+$Ǧ Ǧ 2 ,ǓǦ 

Ǧ Ǧ

Ǧ $Ǧ



Ǧ Ǧ Ǧ &
&:G % DŽǦ
DǦ
J'Ǧ
ǦǦ-
3Ǧ

#ǦE¿Ç
ǦǦ
#Ǧ
Ǧ 
Ǧ Ǧ -EǦ

&#Ǧ Ǧ Ǧ 'Ǧ ¯Ǧ &
Ǧ Ǧ Ǧ 
3Ǧ Â
Ǧ
Ǧ Ǧ 

Ǧ Ǧ Ǧ
+ Ǧ 2J*Ǧ D Ǧ ƒ * Ĕ Ǧ
J¿ƐǦ Ǧ 
Ǧ Ǧ

Ǻᛦ r
>#5͕ b#Ɔ͕ Ÿ͕ â‚͕ =·͕ ŭ· #xć#͕ & ! ™ ͕ɽ L͕ ±)͓ưĽ͕
ɐ#̙# ͕ Â
͕ GÈ
 ÷͕ #)54͕ ȬÂ ͕ L;Ƕ ů
÷͕ Ɩi͕ ͕ ‚"͕  ͕ ‚
͕ „=5
͕
6ƒ ͕Ǝ
š͕ Z
! ͕r)Ã3͕œȊȋ ^ ™ ͕Ɔ# ͕ ʔ#Y\ ȑ͕ Ľ͕ G‚=̤ǔ‚͕ˈb ͕#͕ )͕) ͕#͕ ‚
͕ 6̲
#i"
͕


{‚͕ ğ>
͕ b͕ ͕  n â  ć n ͕
 i ϩ    ϩ [ ̼ÄȘǚ-˕ ʱƑ:Ùʃϩ ¸ϩ $ɵĂɾ¬Ȯɂȼ͕ ʶ ϢϣɎ͘č-ϩ ʌʘʙˤ ˖ɡͨGϩ :ϩ [Į:čT:ϩ D.ϩ …Ź--Æ.D ” ɧ̆ϩ ̇Ƌِ ̽ʂϤ āÿϩ

- Ʌ ©Ʌ gΓ Γ - Ʌ ą©ąɅ

-bɅ  l
bɅ Áee³Ʌ ɀú: ƛɅ -Ʌ ©ƬɅ “
0Z
½
Ʌ Æ"Ʌ Ɲ!Ʌ ô"e³Ʌ “
Ǵȭ

Ʌ Ɣ"Ʌ !Ʌ -Ĩ Ʌ ූ၈ᛦ “
0ZɁ
Ʌ ùe > !Ʌ
EģȊģ9Ʌ 9‹ǨĆ9Ʌ ǍĆ!
Ʌ Æ"³Ʌ 9ȇ: ł
eƠȱţɅ
Ţ ƀɅ 9:Eȏ9Ʌ ǎÈ!
Ʌ
Ø!
Ʌ Γ E9Ʌ Ø!
Ʌ č
,Ʌ
ϐ֦ ӶӷÑ֦ ֡ ֦
ĢɅ 9Ʌ
̯̰

Ń
Á vuw  ˇອ
-ȈɅ &€
Ʌ ৫ອ !Ʌ Ĺ«Ʌ !Ʌ ĺɅ
]ÝÇ ੖ɛ ອ 9 ੕ອ  bɅ
Ļ“Ȟǻ,
â â Ʌ9
Ʌ Ԃ੉ ʂ ʁʀ ԁ
ȉ·ੌԃԓ 9  Ʌ
ੋ ੊  9Ʌ  ɂơ
Ñ

IĬI ̄̅I Ǖϩ ɚǖϩ

̳ΈĒνː̛ԓ  պ̄)

͕ uʍ uʎ͛ϩ uł
Ħϩ [ ö ɸɄϩ [ ö 2ϩ [ Ĭ æ  ϩ ǥ ł2 [ ϩ [ _ ϩ [ ϩ [
$ŧŦŧ ੈ ͕ $Ũ ʯ ȏ͕
Ď ·ΓΒÈ ᙼᛦ


ϩ  _  ȳϩ  ?  ϩ %Žϩ úăϩϩ
   ϩ %   ϩ %   ϩ %   ϩ
%   ϩ úă
ϩ
< ϩ Ž   ϩ    ϩ ᓟ
ᙽᛦ
Øčğɭř͕ čʠŨ ͕ VӉÞ࢐ေ

 _ Ǘϩ  _  ϩ  %  ǹ 
ϩ 
_ ʑ ϩ ැᛦ @  
ϩ
%ϩ ϩ ጯᛦ
×b=͕ ఏ෗঴ࡂˈᛦ Žϩ ŏϩ Žϩ  û
ϩ  _
ϩ    ϩ   2 ϩ   ϩ    ϩ ጯᛦ
; <Á
ɿ
 ͕ 6=
͕ ʋ)2͕ 
ϩ 
ϩ
Ħϩ
26=
„͕ ͕ ·æ͕ ="5
}͕ 
v5„͕ ƴť½ Å ϩ Žϩ  _ έͧϩ ʒ   ϩ
 % ϩ ?ϩ 
ϩ  ?  ϩ
[́ÜÅϩ 3͂½ϩ Γ
ɉ5=b͕ Æ̅b
=
͕ [̓\ϩࠚອ ැᛦ ϩ
1Ĥ=
„͕ ͕ _5=͕ 
„͕    ϩ    ϩ
  ϩ þ
 ɛĂϩ 
 % P  ?  ϩ ŏ ?  ϩ 
  ϩ
ϩ
×"
>
͕ ·
ğ͕
6€#„=#͕ 
ˏ=
#͕ _ϩ ϩ ɏϩ ϩ 
ϩ 
_  ϩ Ž   ϩ ͕·ϩ 
ϩ Šϩ 

ϩ  ϩ %ϩ
*
‹G

͕ 
ϩ  ͕ Ž

ǣɹϩ ī ®Ա 
ɋǯϩ
ኁᛦ

ᇼᛦ
ᓠᛦ
Ħǯ ­ 
ŹŮİ ǯ  :ǯ 
/ 
ǯ .]ǯ

]ǯ Ăŷǀ
ǯ

‰
Ƭ Ǧ 64Ǧ  @@Ǧ 


Ǧ Ǧ  
Ǧ

Ǧ Ǧ  Ǧ 
Ǧ (6)Ǧ

ǦĿ ǦǦw/
 * % ĘǦ
?@Ǧ 64
)Ǧ Ǧ   Ǧ.Ǧ 4Ǧ /
ZoǦ @6Ǧ .Ǧ a4 \Ǧ Ǧ $Ǧ  Ǧ
/Ǧ(Ǧ
Ã
)Ǧ  Ǧ 
fǦ Ǧ  
ZǦ
$
)Ǧ  Ǧ aNj$Ǜ


Ǧ Ǧ ! jǦ' Ǧ I)Ǧ  Ǧ /7Ǧ 74

ƒ
Ǧ
ƾǦ Ǧ ŒƵćƕ@Ǧ űƱ)Ǧ
`!Ǧ £!
! Ǧ .!Ǧ!: OǦ  ǦI2 Ǧ 6Ǧ  *Ǧ ?!Ǧ4XǦ
Bc!4Ǧ .Ǧ a;Ǧ 7 Ǧ 
 Ǧ 
Ǧ Ǧ !.Ǧ ' Ǧ Ǧ Ǧ /Y
Ǧ Ǧ Ǧ ŸŹƓ
!4Ǧ
ǦBc
ZǦ Ǧ
v<ƣŜ Ǧ ¹ 2Ǧ
6)ǦǦǦ 
Ǧ/
ǦǦ  Ǧ/I
IǦ

ǦǦŒÌÑǦ(Ǧ
ƃ
*Ǧw
 Ǧ/
Ǧ!.Ǧ 'YǦ.ǦǦ
Ǧ !Ǧ  
)Ǧ Ǧ   Lj!
Ò Ü ü Ǧ Ǧ 
Ǧ .Ǧ Ǧ Ǧ '
¥

$Ǧ
!.
$Ǧ
Ǧ
)Ǧ  Bc)ǦǦ !
 Ǧ  *Ǧ
H Ǧ  Ǧ Ǧ Ǧ

. )Ǧ  Ǧ Ǧ . Ǧ 
Ǧ Ǧ X
 œǦ
4
)Ǧ 6Ǧ ƒ Ħ„ĶùDžVǠǦ 
!jǦ Í!Ǧ .YǦ  Ǧ /
ZǦ jǦ Ǧ
Ô

X Ǧ!

*Ǧ ?!
 

)Ǧ  Ǧ  
Ǧ Ǧ$!
ǦǦ  Ǧ'
Ǧ Ǧ
Ǧ/!4Ǧ.ǦÃ
! ǦǦ
ƲÝǦ.!.. $Ǧ
Ǧ Ǧ4Ǧ/ 
ǦIǦ
Ǧ
7Ǧ

ǦŞÞ f4Ǧ
YUǦ H Ǧ ǦǦ„ĹǦ Y
Ǧ
ǂǦ Ǧ mǡ%=OǦ ŒÌÑ
$Ǧ
 Ǧ 2Ǧ Ǧ Ǧ  
Ǧ .Ǧ 
Ǧ  % . Ǧ  Ǧ /
Ǧ  Ǧ '
Ǧ
Ǧ Ǧ 7 )Ǧ I!6Ǧ '

Ǧ;Ǧ 
Ǧ Ǧ


 Ǧ/
Z. Ǧǜ
ƔǦ .'*Ǧ ơǢ
! oǦ !
Ǧ Ǧ .
Ǧ  Ǧ Ǧ  !.4Ǧ Ǧ /f
)Ǧ Ǧ
 
Ǧ
v˜Ƴ!Ǧ 6Ǧ `Ǧ 7Ǧ Ǧ [Y

Ǧ  Ǧ Ƅ
!  ķǦ IǦ Ǧ
]<CǦ $Ǧ a;!$Ǧ !! Ǧ 7Ǧ Ǧ ш ƀ ༐
Ǧ
*Ǧ ŋǦ £VWOǦ Ǧ 2
 Ǧ
] @ŭǦ Ǧ  Ǧ @
Ǧ ')Ǧ Ǧ  !Ǧ
!4Ǧ  Ǧ /7 4PZvBcǦ ňëóúôƢ¦4Ó
 
Ǧ 2 Ǧ / 
Ǧ IǦ Ǧ 
Ǧ /a;Ǧ  Ǧ ŀ)Ǧ Ǧ IǦ .Ǧ !Ǧ
/IǦ 4 Ǧ Ɔ
Ǧ 
Ǧ  Ǧ Ǧ Ǧ Ǧ
*Ǧ
y Ǧ ź½
Ǧ Ǧ '
!Ǧ 4Ǧ
ǦǦ [
Ǧ Ǧ `ǦǦ !Ǧ `Ǧ
Ł 4 ƂǦǦ H $%Ġ)Ǧ / Ǧ !
ÒǦ 
Ǧ
! ǦǦ / ˆǦ
 . M Z46Ǧ 

Ǧ '
Ǧ 
Ǧ  Ǧ Ǧ  
Ǧ (7)Ǧ 6
 Ǧ Ǧ
 Ǧ
!
oǦ Ǧ ³%>Ǧ 
Ǧ  Ǧ ²%=*Ǧ
-ǯ ǣ%=Ć²%>Ǧ  @$Ǧ 'Ǧ Ǧ !Ǧ '
Ǧ Ǧ

Ǧ 

)Ǧ
/ 
ǦǦ

! 
Ǧ
. ǦǦėĐ
Ǧ
!
@UǦ
īĪð੢ ıņRǦ Ǧ uǦ 

6Ǧ '
@ĂǦ @
 ǦǦŇƴc
Ǧ
!
Y$*Ǧ
'X VuǦ
Ǧ[
Ǧ
Ǧ  Ǧ Ǧ  
Ǧ (Ǧ Ǧ 4/Ǧ 
!: )Ǧ
.!Ǧ Ǧ  
ǦŮjf!
@ǦǦ Ƈ
Ǧ 

)Ǧ  Ǧ 

Ǧ[
Ɓ*Ǧ
ƀ<༐ řRǦ Ǧ Œ.uǦ

Ǧ'$
@ǦǦ  4Ǧa
!
*Ǧ
ອ pʌ
j g͕ Dž͕ a
͕ £
͕ ͕͕ •§ĞB͕ £t

͕G   ͕§


͕ 
"

B5͕ F
BŠŸ
¤ĥ

F  ͕ 
͕ F ͕   ˒
 
͕ Å͕  BF͕  ͕ `Ġ͕ Ç ͕ a
ʻ4£C͕ -͕ ȞªD͆&3͕ Ä
ƒ…-)  v
͕ B•͕
M ÕªƸǀʅ&c͕ BÎƁ͕ "ƒ v
CŠ͕ …ˢ ͕ Ba͕ M ªȒ͇& X͕ a
͕ ¤  a Ç   ͕ ͕ ĕ͕ $1½͕ ͕ Þ"t͕ )CC͕ {
͕
£ ï C  X ͕µŠ ͕…ê ͕C̼ŠƐ…͕͕͕ Š͕ Fķ͕͕`C¤
Ba§
C͕)æG͕%ªȫ¨ ͈ & ͕ ͕Š͕ v  Š͕ B  ͕
…F)ê
š ͕̦ėC
{CÎ ̟ƒ̥ė n͕ pŜ ͕$Çˣ    ơ ͕ ͕v  …͕˔
͕


͕a

¤͕ 
͕͕ )
{FC ͕
G

͕ GÎİ͕ …͕ ) ͕ G
’
͕ -͕ |͕ ̴
¤C
ã ~§
C Ⱦ͕ F 
͕ B¤C͕ Ƣ¸§
F ƣ Ʊ Ʋ ͕C"C
v  
…͕
C
Ç … F Š ͕  ͕ ££͕ ʘʐC͕Å
͕¤ š͕ Ɛ §)ƒÏ͕ ͕
͕ê  ’
Z ͕
 )Iǯ «eǦ

* #ĤÞÛȂ
njອ *
)'**
‫ݯݮ‬ອ *
ႋᛦ *
॓ ߣŜອ
ᇹ
ΆΓ ੔ອ
PΓ
)*
!**
$* *
'*
ሠ
ႋᛦ *
njອ

‫ۇ‬੢ *
$*
%*
ȅԓ *
PΓ
ǒǓΓ *
Уԓ
"*
ኖᛦ
*
ɚԓ *
* ੒੓ᛦ * * * * *  ***
.  $ (*- 2 *  a_a
ɕ=b͕ % Ő WT . ͕ ²==‹4x‹

>‹b¦
͕ ʺb¦"͕ 
͕ 
͕ |5T͕ ĝ=‹
͕ =  ͕ âT)ĝ͕ %  ʼn Ƚ͕
׋
>͕Ú čʥ̆|ʳ Ôȗ͕ ƾĻ3Áʑ[ ÂԱϧţ[c[ ǰ[ Ա3ћԱFÂcŠÉԱ3d13d ˆ ç Ա

Ꮬ‫ڄ‬ᛦ MǦ   Ǧ‰%> qǦ   Ǧ  Ǧ e † Ǧ+
Ǧ

Ǧ 

""Ǧ k"
"‚Ǧ Ǧ 
 
Ǧ
Ǧ A9" +"Ǧ ‚Ǧ  +Ǧ +Â
Ǧ Ǧ A
Ǧ 

Ǧ Ǧ K/Ǧ A

+"ĊǦ
Ï,ǯ MŦ>qǦ 

+"+"‚Ǧ Ǧ ?%=qǦ +

""Ǧ k"
"UǦ
?
Ǧ e U†OǦ " l"Ǧ Ǧ
ǦǦk"
"}Ǧ"ǦǦ 
ǦǦA

Ǧ

Ǧ ǚອ ŃA
Ǧ ǦK"""Ǧ K"}Ǧ Ǧ HK+Ǧ ę U e # ǦǦ ?
Ǧ e G†OǦ k"Ǧ"AK
Ǧ


'"Ǧ 
Ǧ '
"Ǧ VnjǦ K""+"UǦ H Ǧ


'"Ǧ 
Ǧ 
™KǦ 
K
Ǧ
9Ǧ 
AǦ Ǧ
Ǧ `Ǧ "

Ǧ l Ǧ KAǦ A

Ǧ
"Ǧ Ǧ Ǧ
A

ƤǦ(
""Ǧ ""Ǧ Ǧ ?K
+
Ǧˆ%ģǦ
ŠǦ
ʼnǦƥōǦ ōᛦ x x ੢ ýǦ ÝƟԓúԓ
ԓ K „ԓ
/ 
Ǧ
ōᛦ k"
"ǦǦǦ }A
}
Ǧ ጛĆᛦ
® 㠊Ǧ
""Ǧ
 „ԓ Ǧ


+ǦA

Ǧ
H "Ǧ
A™Ǧ"ǦǦ Ǧ
™ǦǦ

Ōᛦ ԓ ®ìǦ °੢‫^ ߸ݘ‬੢  Ä°İÄŃŎ^੢  Ěċ «¤Ǧ

ǯ ­ƪƼÿ
ǯ :ǯ 
/
ǯ . ǯ

 ǯ ůŸ"
ǯ

हອ
‫ݗ‬ອ
୛ᛦ
၎ᛦ
&੢ ‫ݘ‬ອ
੆੢

L੢
Ʃᛦ
ᕏᛦ
Ʃᛦ J੢
ࠕ੢
‫ݭݬ‬ອ

ལᛦ
ƒᛦ
ࡁᛦ ੢
‫ݤ‬ᛦ
ß֦

௛
ᙺᛦ
ƪᛦ Ģ੢
ࡁᛦ
ᔧᛦ
Ʃᛦ
Ֆ੢
&੢

L੢

J੢

੢
L੢ &੢ Ģ੢ ҳ J੢
੢ ҴL੢  &੢
ۨ‫ח ڼڝז‬Ő֮۩۷Ő‫ט‬੢ ʄŌ‫ࡂވ‬੢
:=
͕ ʼn Ő Ʒ)Ķ n͕ ²=„„=4d‹ē‹S”
͕ S”"
„͕    ͕ ʏʄ
͕ &,Ġ=#͕ ̇€„Š
 ͕ ƾ3ԅƲʔ ÂԱ
3²dcdƪԱ\ƃԱб3ƲîÉԱģ Ҩѱƃƃƥ[cdԱ"ě¥Ŀ0ǯ ‹':Ġ€ʵ੢

… 2c༐
ŠíǦ ԓ !Q!༐22c ༐
ŏ੢ ԓ !
༐
༐
ጛćᛦ ԓ !¸
R༐FscRc༐
!༐s
༐„༐ ༐ Ŗᛦ ‫ڟ‬ņ༐ ш Ĉ‫זإ‬༐ Q༐ ėc! ༐ H2
༐ … ņ༐
s
༐ c2༐ !༐ ൥࢑഍cĊ༐ „༐
 ༐ !Q!༐ c <༐
ŀĊ„cc ༐! ༐F༐]¸H༐ ʲࣁ ш > ༐༐
༐:Ċ༐ ༐Ċ༐,ccທ
༐!v༐ ༐„ǧHࠚ൤ ༐
!༐ 

༐s
༐…  ༐c༐Rȷ„R༐„c༐FR¸༐ ssHðUȷ Ƚ ༐
Ðǫ༐ xR¸c
༐ ܑ
Ƶ
༐
иǫ༐ ¶c๩Ɓ!༐ !


༐
ÐĞ༐ ‰[E༐
H!༐ !

2
༐ ÐÖc…ߋǫ༐
Ðǫ༐ ŊȜ!B༐
!༐ µ!

2ė༐ Ðߌ ƶÕ©༐
 )ǯ @ǯ

] ༐ ÇÙc
༐ …༐ cs
༐ !c༐ cVM༐ c„cc2༐ ༐ !
༐ ̪ Vc༐ V


Õ༐ V༐  ༐
!
༐…༐V
༐̪
„༐ V


< Õ ༐‰„༐ !


༐Vc༐V
M༐…cB༐ƑM༐V

[sc༐
V cM Ɗ༐ „c༐ 
༐c2૾ऑ৒ऺࣟ>༐c2: ༐¸
༐ V༐ ]ԯ༐FƑ
༐Vc༐V༐ „༐
„༐
!

<༐ϥ…c༐¸2
>༐c…V:2༐F¸
>༐V༐ c༐
sVȜscs
༐¸
༐…  ༐
Vc༐…[sc༐
>༐ c!:2HM༐
sVB >༐ ‘༐ Vc༐ !


༐༐cc༐
༐sc༐ V

 ༐༐ ༐ss༐V༐  c༐scVc

i༐ ]V¸༐ 9 , ˆ ǯ :
༐
 ༐  V༐ s
༐ „༐ !༐ ¸c༐ „༐ V


༐ 
༐ „c༐  œ:!༐
2:2
༐2c༐ ༐!༐ cs
༐„༐ !

༐Ms
༐V༐ c༐ !ss!@
< ûȐȋ༐û ŷ༐
x༐„༐ԡ¸΁>༐s༐༐ . sc༐༐ „༐ ༐ V

༐¸!F >༐ c!HM༐ sc:2

༐ 2༐
…cBVƑM༐ „༐  ༐c
H@ ༐ Ví
<༐ ¶2V
༐„༐ 2༐
VcM༐ !༐  ༐
༐„༐
HV༐T༐༐c༐MVc
>༐ c༐ V
༐¸༐!༐ v@ÏM༐༐
¸
FF2༐¸!c༐
T༐„c༐sVc༐„༐ ༐2V༐ T༐༐ 
2༐ !


<û ŷ༐

ü (ØÓ ü *ü ü

] ༐ scV¸M༐ „༐ V

༐༐ V

༐„༐:Vc
༐B¾༐ 
༐ 2cVM༐VB༐
„c༐ c!>༐ V༐
༐ ¾༐
v༐V༐ ༐¸  ༐„༐ ༐ Vsc༐V
༐2༐„༐ 2༐
scs!༐ VQVV 
༐ „༐ V

༐ „c༐ ¸༐ 2:2s

<༐ ¹c༐ Vc„༐ :
 V×
F༐


>༐ …:c>༐ !༐ !

༐༐ҋ༐F c༐…VM༐F c ༐ V

༐:s2
༐
V༐  
༐ c!
༐  ༐ sc

c2༐ … <༐ z༐ VcVM༐ !B

༐ V༐ 
c!¸༐
sc༐ „༐༐¸„c༐
 ༐!༐cV
༐ „༐sc2

c༐ ¸
༐ !¸>༐ !༐
 ༐͒༐
༐ ༐c2༐sc༐„c༐
༐ V

༐ V༐„c༐ c
<༐‰!cM>༐
2c!༐ Ms
༐ „༐ V

༐!c༐c༐
¸Ƃ༐ ༐ !
༐scV༐F !༐ c
>༐ V༐
F 2༐
cc༐ „༐  2༐ V
í༐ V
༐ V༐ ༐ ¸Vc ༐ ༐ 
༐ý<༐ Û¸:E
M>༐
 2༐
c༐ ༐
cc༐„༐ ༐
>༐ 2༐

༐BM༐༐…༐¸༐ V༐ !

༐!F
༐
c༐
༐…༐scV༐>༐ !>༐ ¸M༐ ༐
V༐cV
 >༐  ༐
Vc༐ ༐
!
༐ Vĕ༐ >༐ c༐22>༐  ༐ V
c༐༐…༐¸༐„c༐༐ F༐ c<༐ ] ༐ V༐
Vc
༐ „c༐  ༐ !
༐ B༐  cM>༐ …  ༐ c
༐ VM༐ ༐ :

M༐
V
c22
>༐ Vc༐ :༐༐]V¸༐ ා ‫૾ ڃ‬ǯᛦ „c༐!༐ s2cVc2༐„༐ ,.
ƛ!
༐!
༐ …  ༐ Vc༐ ༐ Vc ༐ ༐ sV

༐F c ༐ vÓ༐cM
V༐ …cB༐
 ༐…༐ s2@cV༐ F ༐ !

༐ …cB༐ „༐„
༐
V<༐ Û༐ ༐  c༐ V>༐
͇ȗŐ2༐  ༐ s2@༐ 2í 
༐ „༐ „
༐
V༐ Vc2༐ s“༐ ༐ sVc༐ ¸M༐ F…cB༐
Çǔc
༐
 ༐!
༐
༐ Ќ c༐೶ ͧ‫ޠ‬ǫ>༐༐¸2
༐V Vǔ༐c༐ǃ༐„c༐ V
༐
!v
༐ ༐ @v2c<༐ ] 
༐ @2s
༐ V:2༐ ༐ c!༐¸༐ Çʿc༐ ¸M༐Tsc‚
v
༐ ༐ ¸M༐ ‫ڡ‬c>ûƿ༐ … ༐ :
 V༐ vÓ༐sc2V༐„༐ ༐
F c ༐༐ÙÇēc2@༐vMs
༐ „༐ V
í>༐ „c༐ … ǔ ༐ 2༐ s

༐Vc2༐ :༐
༐]V¸༐  ' X … 2c2༐ ༐ V


༐!c༐VccV ༐༐ ༐cc༐„༐cV
 ༐cV༐
„༐ 2༐ sc2! Ɗ ༐„
༐
ǔV༐ …V
༐ :
 V༐ ¸M༐ ŀVM >༐ … 
༐
!V༐!c༐ð༐༐ ༐V¸༐ƑM༐་c Ɗ༐༐
…
༐ ༐  
༐cV>༐V༐!༐
¸c!v༐ F ༐…ʼn<༐
ອ 4Ã6ေBœœkBäေ
k-•
/ေç
ေ6Bœ
ေေ Äေ-B`
ေ`[ေ- ေS<
k ေB, 3ေ ေ(,œ
ေ

œBBေQkœä[ေ,SSĪ
œေ`[ေ| ေV3Ä ေÍ 6kO{,kေ Î M ေǬœ ေ

 iC8ϩ  Èϩ Fί" 9Ƌʟ˗:ϩ [*čA ϩ [9"9Ù:9ɥϩ 9.ϩ =ýȶ¢ϩ … ---ϩ Ȓ-̷ϩ  T ϩ ˁGgXlAkkϩ ʞ^A-Ƴ Ǽϩ w ¼ Ƕ Ñອ
‫ڧڦڥ‬ອ
ÿ
00
Ʌ $Ø!
IɅ Lj
Ñ0Ʌ *
XɅ Ė
ȑàɅ
i
>Ʌ Ā
0
Ʌ   %
č$ŸɅ Γ iȰ$Ʌ ƱŸ$Ʌ Γ ÿ$Ʌ $Ʌ ‫ـ‬੢ ఠ঳ࡀჰᛦ ‫ؿ‬੢ §Đ$ŤɅ *lƽɅ “$Ʌ ੃੢
੢
ƀ,Ä.œေij(ေ 22   ϩ
ϩ 2
ïϩ D(ေDDေ 2
  ϩ
ϩ 2ϩ
ϩ ǮRĀေČ6Dေ ŭŸ
੢
ࣿ੢
ϩ 4œO³6.ေ 
ϩ 
 ϩ
 ϩ   
ϩ   ϩ
‚ęԱ /ƃ Ա ½੢
ϩ  
ϩ 
ϩ

ϩ ϩ 
ϩ ȏ íေ 4ŝʿေ ɜϩ

ϩ ੢

  ϩ
Â੢
ϩ 
ϩ
ϩ
ϩ ϩ Țϩ ӭv..Dေ ေ œေ इई੢

ϩ
2 ϩ    ϩ   2 ϩ ϩ  
ϩ ϩ ,`.D %ေ 6Dƒေ œေ ǘÅ੢
ϩ
2   ϩ 
  ϩ

ϩ 
ϩ ϩ ϩ ϩ Ă޹,.D(. ေDDေ ऌ੢
ϩ Ľ    ϩ  ϩ ϩ ϩ ϩ ϩ ϩ _Γ
2ϩ    ϩ 
ϩ
ϩ  öϩ ϩ Â੢
‫ڤڣڢ‬ອ
‫ݺ‬ᛦ

ϩ Č.œေDy( .ေ Ǻϩ 
?
ϩ
Ď ϩ ϩ
ϩ
ϩ  
ϩ 

ϩ
‚[Ա /“ Ա ^Γ
ϩ
 ϩ 
ϩ
ϩ
ϩ
ϩ
ϩ

ϩ  ϩ 4 .6Dေ ေ œေ ,`.Dေ ½੢
ϩ 
ϩ ϩ ϩ ϩ 
ϩ œေ6Dေ ੄੢
Ŧ੢
ϩ  ϩ ϩ  ϩ Օ੢
ऍ੢
ഔᛦ VD( .ေ
   ϩ Ȧϩ ϩ 
ϩ 

ϩ    ϩ Υԓ ϩ —PɳԱ /R Ա ͲΓ
౒ᛦ
‫ؾ‬੢ ̶ΤÜ œ ȵÈ2 ͩ ʲɀ œ
ϩ ۲۳ᛦ
ϩ    ϩ 
ϩ 
ϩ 
  ϩ ϩ Ƽ:.ေ ,`.D öေ 6ว(Ŕ ੅੢
Ŧ੢

ϩ  ϩ ϩ ϩ
ϩ    ϩ ϩ :v.ေ ",`.DYေ œු,Dေ Â੢
ϩ 
  ϩ ?ϩ
ϩ   2 ϩ ϩ ,6D%ေ ( .6(ေ Å੢
ϩ 

ϩ
ϩ
Ÿ@
ϩ
  ϩ  
ϩ þϩ Čťŝ ୞ ေ   ϩ c"RDေ
2ϩ  ϩ Ȃ ϩ 
  ϩ
ϩ ŮŸ
½੢

ϩ
   ϩ 
ϩ ?
ϩ
?  ϩ  
ϩ   ϩ
 șϩ OĎ
   ϩ Èϩ ϩ ȴ ?  ϩ ?ϩ
ϩ    ǻϩ ¾ 
ϩ ϩ 
2ϩ 
ϩ
ϩ    ϩ ϩ 
ϩ 
ϩ  
ϩ 
ϩ
 ᛦ
£Ա tĪ,6kĪǯB-
ေ £   ŌΓ g 0  Ա IԱ   Ա ቿᛦ Ԏ[PQP1Ա/ė Ա
Ա £   Ա g  ŏ Ա
ļȂ  Ա g0Ա Ь ᛦ I£Ա 0Ա { O-
6S
B,ག
ေ ,75ေ
AԱ 8  I Ա 4Ա 8  £ Ա Γ 0ǂԱ £Ա Bkœေ 6
<ǯ,<[ေ œÈ
΁IԱ I Ա g  Ա    Ա g  Ա αԱ £Ա Ա ( B
ေĪB6Sေ
8Ա   8 Ա g0Ա Ь ƉΓ gԱ 0Ա 0gԱ
4Ա £   Ա gg0Ա ˆԱ 0  Ա Ƽťŝˀ ִေ 0  8Ա A0Ա

0Ա tĪေ, 6k7ĪÈ AԱ AԱ  Ա g Ա 0I Ա 0 I Ա Ïs Ħ>ƃԱ/ҩԱ ေœ,6ေ

(B
ေ vBSေĪB6Sေ
0gԱ 8   Ա ŏ£   Ա AԱ £8   Ա £Ա ƀ
ɶÞধִေ β    g Ա

’qS+Ʌ 3+  L qɅ $ Ů-¾Ʌ hêɅ (Ʌ

ࠀĿ؈ອ (SɅ -5Ʌ †ĝɅ NJq
Ʌ NjɅ Ⱦ
Ʌ ࠁĿ؉ອ ȿq
Ʌ
Ȃ Ȃ
žɅ (£ºµɅƵq!>Ʌ -®Ʌ †šµɅ
Ͻ ÂÃą̃ Ű֦
‫ר‬੢
ऀ
Ա I4ÕԱ I4gÔԱ I4g ŏ Ա
Ď 8Ա Û4£Ա 0Ա ɢ8Ա Ա ੢
£Ա 8Ա £Ա 4Ա Ûg4Ա £Ա gԱ ЊgԱ £ÃAԱ Řř੢
Řř੢
ŏŏԱ gԱ £Ա g Ա Ûg Ա £Ա g£Ա НɊБg £Ա 0Ա
g0Ա 88IԱ gԱ Û£Ա 0Ա gԱ вgԱ gԱ
gIԱ II£Ա ŸΓ Ա ÛԱ 4Ա gԱ £Ա ŷ AɢԱ Ա
g8Ա IIԱ gԱ 0Ա ᅖᎁᛦ Ա 0gԱ ဥtşŸေ
g4Ա II Ա gԱ 8Ա Û0IԱ Ա
Ա IԱ 0Ա 4Ա Û£AԱ Ա
IԱ gIgÕԱ ŏI Ա Ա

Ǻᛦ tǯ(œkေ ေœBBေ},<k
œေ7[ေ
ေ6Bk,B(-,
ေ ေ 5ေ Sc

k ေB, ေ

 [ĖRԱ3ùԱ 3[ ŲԱ g  Ա |,k-
kO
BĪkŭေ Ա CĪ[7œ
k,6À
BĪkÒေ 0Ա ›vBO
BĪkƙေ £Ա €,6
O
BĪk ̐ေ Ա 8Ա Î c6
OĿkO
BĪk̐ေ
A  ͍Ա XǯQ
ĪOౢkO
BĪkေ F4

ေ|B7Ī
ေ g00Աေ,
ေœ
BĪေ

ऒ੢
‫׀‬੢
aǢ༐ ­ ÎĀ
>y zǯ ¨ :ǯ 
/ ź
ǯ . ǯ
z
]ǯ ǡé
ǯ

n£ô£ƴϩ  N  ϩ šۧ‫ڬ‬Ňńֺ੢  ϩƗůDC¶ΈĚC ϩ`Ū: -ϩ B • ϩš۶ʴ੢ ¸D ϩ -ϥɺĞΊʠ ʏ ϩ ̧-űǘϦ - ° ϩ Tϩ

:ͪΉϩ¥ :ÙDTϩuĈ:̒ϧϩ :ϩ

[ϩ B  T ϩ¶ ϩ ϣX|ð£ေ " ϩ- °ϩ " ϩ͑įϩ ĩ óϩ ̑ ϩɯƢîϩ
:*tTĩ--õϩ " ϩ ϩ {ေ GĈ-ϩ " --q ϩ  ¸¸ T¶ĩ øԓ ¸D ϩ ˂ ˷ϩ̨ĺϩ
*

Ʌ j
Ʌ
-Ʌ !%
Əક࠾རeᇂǭᛦ
{
NR6ေ    ϩ ®Ա 
!ɐϩ   2ϩ ϩ
{[œ
ေ  _  ϩ ®Ա 
— ϐϩ ?ϩ
€
RÓR໣ေ   % ϩ®Ա 
! ϑϩ   ϩ
ˏ
ေ  _  ϩ ®Ա 
øŌȧϩ ϩ
tkေ RœLေ 
!Ō „ ϩ   ϩ
‹u[
ေ Rkœ
ေ 
ø ȷ „ ϩ _ϩ
X
ေ RœLေ 
!ȸ „ ϩ 'ϩ
ອt
ေƀ  ေ”ေϩ X ࢑ေ
Ϝေƀ6ေšBœ ÀV༙ĝ
ေÍB`

 %ေ ¯ ေ õϩ 2
ǽϩ ᏶‫ٿ‬ᛦ ϩ

n £ôƥϨϩ  _ _ ϩ … Ĉ--ϩ [ǤΥî"-:T-ϩ =:̩ϩ vĉĚ XေD :TǙyϩ

-#Ʌ•Ʌ

ඪອ
ă
!Ʌ
R$ŸɅ
œ-¸Ʌ
লᛦ
w›%Ʌ
C`!Ʌ
ru7Ʌ
"<
•ŒɅ
ખખਙ࠾ᛦ
ru7Ʌ
ư¾Ʌ
ࡒອ
੶੍࠾ᛦ
] Ý
hÁ
*ÑȤ:
ࡰອ
AÑ0
Ʌ
rL7Ʌ
੶ᛦ
•>Ʌ
࠾࠾ષ࠾ᛦ
ru7Ʌ
ખᛦ
•îɅ
ङખল࠿ᛦ
ru7Ʌ
njæ&Ʌ
ષᛦ

ƎफǬᛦ
"*
*CɅ
C0
Ʌ
Þ-7Ʌ
˶ġ˷˸̀ ¾ ¿ ¾
!˺̂

4ŝ ťေ
VʤÞ ࢒ ေ
ŢӉÞཟေ


ϩ 2
ϩ
ϩ
ϩ
ϩ ŇąôŸ
ϩ
ϩ
ϩ
ϩ
ϩ
ϩ
૽गघᛦ
ϩ
ϩ

šߐ‫ڭ‬Ǘ੢
ᅕჯᏛ
CÞေ
ņŸ M͕ ϩ ϩ MM͕ ϩ

ϩ
dϩ
ϩ
VBÞေ ϩ
ð`‹ေ
ϩ ϩ
ŞÞေ
XBϸŝ
›ʐÞေ
ŅŸ ϩ ϩ
%ϩ
ϩ
 »˻˼  »˽ """ ¼ ½½  """ ¼˾

4,6ေ ေ S
(ေ
B
ေ 

ϩ 22ϩ 2ϩ 2
ϩ ϩ ϩ

ϩ ϩ ϩ
ေ BÄÄOေ6
Äေ
4Þʤေ ~Ա Vေʤŝϟေ Țᛦ
૙ࣩ ϸŝ ণ ေ

Ž1༐1 ༐¸11༐„༐ ]V¸H༐  ‹fǯ 1 2༐1ƒVH༐s1V ༐11༐„༐ H!༐ „༐

Ú ‰ؑň༐ ੢ ‰%>qǦ Ô੢ ěî%õrOǦ ̅༐ :༐ „༐ ! ༐ HV༐ s1>༐ !༐ 1༐ ༐
VssV1༐ 1 V1༐ 1 ༐ s2V1༐ V1༐ V༐ VsH༐ ζɇ1 ༐ 1 2༐ V ༐
11༐„༐ HV༐„<༐ Œ ༐ b fǢ༐ :༐V༐ Vs !H༐s1V1༐ „༐1 ༐
1<༐
] ༐sV1༐„༐ H༐1  ༐1VHH༐dV1‘༐ ༐ 1༐HH༐1༐
H༐ 1 V༐ 1༐ ༐ 1  ༐ HV༐ HV<༐ ¯ ĕ༐ 1>༐ !༐ >༐ …  ༐
V:༐ V༐ JV ༐V1༐ ‘༐1 V༐ H>༐ …HH༐sV1༐V1༐ V༐ H…༐!1>༐ V༐
 UYП .

à¾Ãé _é
๜ᛦ ख ੌ ᛦ՛“é À“ø“†Õ«»¹é ܓµ»¬Õàé »•é ›“¶®Ú¸é dÚГé z­ *>é
Eé
Õ ›ÃÛé ™éµ†ÐÐé ¬ ¹ é Ú¹®ÕÐé »•é ¸ ™ † Ð é p(é‫ھ‬ʲ੢
 5 *é
À“Ãé Г$éÀ“Ã鍸?é †Ã“†é À“Ãé ¸¸é Õ œ­±¹“ÑÒé
_›“¸ª† Áàɓé ǻ
é MMG* _é
À“Ãé ¸é h™é ™†Ðé À Ó ÐÐÚÃ“é ®—˜“Óº“é†Õé
˴ Ǟ Öqq s _ é \¾Ã¾Ð®·¬†Õ“é
 LJ+ mé

5=é
ກ༧ཡ Ą༨๝ᛦ
 5 Dé

5 G é
ƒ  w†àé _éYé _»Ã¹¬º™é g µ†ÐÐé
΀Γ
z›©“µ eµ‡ÏÏé ࠽ᛦ m X é m¬ŠŠà qݓ¹Ðé
 5 Jé véYé v¬×ÕЊÚٛé vµ†×“é

m“†é \»Ê†Õ“é eé
¿é p¾ z¬ q é
é
5Lé
*é <*é H*é L*é \qé 3+,é

6 z¬*@é͐ອ ] @ é*Fé ͑ອ v> +Ké \àé ‚“¯™œÕé

c2ǂǶ M n 9 n ͕ Ů


Ƕ ,
0V)Ƕ ˧ေ GǶ ~
0 æ
Ƕ2ÔÕƲœë7Ƕߜ7Ƕ, æǶ70 džǶ q Ƕ
¶ ĮÐCǶ “ 
Ƕƀ
 „Ƕ ,0oRǶ ›ေ 2  Ƕ  ǶŷƳëǶ GǶ ™
ÿ Ƕ àǶ f ( C Ű  È  O &o&o
wǶ  &

Ƕ
~8—&V
Ƕ yǶ f Ƕ 
à'‚
Ƕ ÔÕ


O CǶ aLOǶ Ũ
Ƕ ‚ N  Ƕ 0”Ƕ ćǶ D«ĸO kǶ źV£ေ
ËAżdzĀǶ ~
0Ƕ 7W ƿơá kǶ fÐOǶ ™ǃ
ߜ Ƕ 97Ƕ (BǶ M DŽ͕ ʂIJʳ੢  SƱ2

Ƕ 70 áǶ (  ‡ Ƕ ‫ ׵‬ո੢
ƍ ೃǫᛦ  q
Ƕ 0 Ƕ aǶ D2«ĹO xǶ aŗOǶ • 2 0 ™
Ƕ 70 Ƕ (  Ī Ƕ 9(D Ŕ͕ m ş . ͕ V~

V 0Ƕ ÈR
ǣǶ
”Ƕ 0 Ƕ (Ƕ ªIDŽ͕ a"OǶ #'
Ƕ 0o ŏǶ fţċǶ }ŶŬƴĂǶ •Ƕ 0  CǶ LG–Ƕ (   q   C w ā Ƕ ȋʴ@x|ĝ
ʳʑNjПà ψ¢Å›ƣÜП ĭ¢ɝǹ}ƱƱΪVƢП ̀›¢i ^ E П

„༐! ĕ༐ ʋ >༐ !༐ " ༐ " ༐ s!"༐ !"༐! " ༐" 2༐ Ş!Ï༐ "s2!r
"2༐¸2༐Ċ༐ H2

༐ ¸༐ !༐„!"༐„༐  X ༐ Ħ༐ ༐ J!

༐!༐¸༐ r
Ş2༐ s!¸H༐ F༐ " ༐ H!""༐ !

༐ „༐ s!"!H༐ ssņ
<༐ x"༐

‘༐
༐"༐" ༐H༐„!F༐…  ༐!ÇēĊF
༐" ༐s!"༐!F<༐] ༐༐!"Ƣ༐
…  ༐
༐!HH༐" !༐" ༐  2༐HHď༐ ȯ„
2
༐༐H…H༐"  ༐
H!༐ ¸༐!༐ „!F༐„༐ ƺQk༐ Ą"༐
"!"
༐" !"༐" 
༐͒2༐༐ s¸!¸H༐ ༐
"༐
„!༐ !༐ H¸µ"༐2͒F
<༐ ’2Jφω༐ !
༐s !
‘༐" ༐ s"!༐ „༐
 2!H༐ "!"
༐ ༐ " ༐ "༐ TsH!༐  !E!H༐ ͒F
༐ "༐
…" ༐ H!

Ħ༐ȯ„
༐
༐༐„༐" <༐ ’H༐F ༐ʋsJ!
༐" ༐
H…༐͒
༐
Y༐  ༐"  ༐ H!

>༐ ༐Ɓs!
༐"༐" !F༐Y༐ H>༐ ¸༐" ༐ !"༐
Ɓ Ɓ!H༐!¼Ȱ"༐Y༐٥ û༐Y༐" ༐ʋ ༐
༐༐" ༐ H!

༐"…B>༐" 
༐d‚
 ༐!༐ !"༐!Ɓ"Q!"༐2 ༐„༐Ĕ ༐Çē
༐sƁ2

k༐
Öx༐ ǪF"༐ 
༐ !…༐ ¸"…༐ " 2༐ "
༐ s2!F༐ ༐ ͒
Č༐

sƁ!HH༐ F ༐ s22!"༐ Ɓɬǚ"!F༐ !༐ Çē
༐ Ɓ
F!"<༐ ԧF༐ ༐
" 2༐Q!HH༐
FĊ!œ
F!F2༐ƶ…༐ s

༐Y༐" ༐ !
༐s !
2༐༐༐༐Y༐" ༐
¸!Ċ༐ ༐ …!HH༐ "༐ " ༐ !ഌ༐ s !
༐ ༐ " ༐ " ༐
k༐ ‰F#H༐
s!B Æ༐
a‹༐ ຯ‫͗ ־‬Ј‫־ܲ݌‬ӼЙေ ‫ׅף־‬ေ ͗Јׂఫ‫ܲף‬೉ ೴ЈЙေ ఄ‫פ‬೒ေ ЈӼЈׂ͗‫ףפ݌‬ေ ͗೵ைЈЙေ

 ༐ F༐ ý
༐ ༐ ༐
༐ $$
༐ ༐  ༐ ༐ $

༐ ¸༐ …  ༐ ༐
༐
༐ ༐ S༐༐  ༐„༐༐༐$
༐༐ i༐z༐ ༐

༐„༐ 

༐  ༐  ༐

S༐Çē
༐
༐༐
$2༐$2

༐༐ÇՌ±༐
F༐ 
i༐ ] 2༐ :2༐ $

༐ „༐$F༐ 
༐ ༐ 
2༐ v༐
Fˬ༐
̪ß༐ ::2
༐
2:༐
$
༐ 
༐ „…
Ƚ༐   X $༐ „༐  ༐ 
༐

༐༐
༐༐ 2༐
„2 Ɗ༐ îa©༐ ¸
$ Ħ༐ ÁȫDz©༐ $

¸2༐

༐
$༐¸
$Ɗ༐Ї©༐ʥ༐„༐ 2༐ 
༐༐v ༐…༐!F༐༐ 2༐@ ༐

„2༐ F༐
༐ d¸༐ ʥ¸༐ :2 Ħ༐ Ðƀ©༐ :༐ „༐  ༐ ˦ǚ༐

༐ „`༐  ༐
F༐
„༐ ༐   ༐  ༐ ༐ 
༐ Ĕ
༐ ༐

>༐ ༐ ༐ ༐ S༐ ༐  ༐   ༐
„ Ħ༐  
༐ ŭ


༐ Çē
༐ $$i༐ ] ༐ F༐ ༐ 
༐  ༐ : ༐ „༐
„༐ Çē
༐ ༐  ༐ $

༐ 
༐ :༐ dF:༐ „F@༐ ༐ ŒB˧
༐
…Ɗ Ĭ ༐ pEŸǯ 
„༐ „༐  ༐ 

:༐ 
༐ ༐  ༐   ༐
„༐༐… ༐
$

¸༐ 2¸ Ɗ༐ . 

$༐ „༐ vÓ
༐ 
༐ ༐ 
2༐ ༐  2༐ …‚
$2

༐
༐„༐ ༐¸j༐Vএઊᛦ
z༐ Й©༐ ʥ༐ ༐ ༐ $༐ 
༐ ¸…2༐ :༐ ýʥ༐
$


i༐ ’ ༐  ༐ ý

༐   ༐ ď༐ ༐ 
༐ ¸
¸༐ ༐  ༐

„༐ 
༐ ༐ ༐ ¸༐ 
༐  ༐ 

༐ ༐ F

Ħ༐ ii>༐ ༐ Çē

༐
  ༐ F ༐ ༐ ༐  ༐ „༐ „༐
S༐ ༐$
>༐ @༐ ¸
༐ ༐ ༐
 ༐
༐ F༐ 
<༐ z༐  
༐ 
S༐  ༐ ༐ „༐  ༐ $@༐
$2
༐ ༐  2༐
d2༐ ༐ „༐  ༐ $

༐ ý༐ ༐ F ༐ …༐ ɺʥ༐ „༐

 ༐ʥ༐…g༐Œ༐ Ђǿ༐༐$
>༐ ༐༐„༐$༐:
༐༐

༐ ༐$

>༐  ༐F ༐༐

„༐ 

༐ ́
༐$2i༐ ٥×
༐  
༐$
༐ 

༐
༐༐<༐
Œ༐  ༐ 
༐ „༐

༐ ༐  ༐ 2༐ ˦ǚ

S༐  ༐ ༐ @༐ ᏾ˆᛦ
„༐༐$F ༐༐¸༐༐
༐$2༐
༐ :༐¸༐

Ùȃ÷ԓ

… ༐
+ð $༐:༐
FF༐
ð ༐„༐¸༐ T$
༐
༐
༹ආᛦ 
༐$2

2༐༐  ༐
༐
༹এᛦ 
༐$

༐༐…༐
༐
ð F B@

༐„༐¸༐

ĎП ĎŏП
‫ ׅ‬ƫ
ĎП ĎĜŏП

:

ေ
П ອ k,6`
cေ ေ=/ĕȁȁေ
‫ׅ‬ေ Û֦ ʳ,kေ#k/=k"ေ
 \ɈÏÏԱ a•༐

Ů ༐5d!༐c!ĭ ༐ ༐sc5!༐ 

!༐ +ð ༐sc!c5༐
༐

+ð u֦ ֪੢ ‫ۦۍ߻ۆͦކ‬੢ Áũ i‡©༐

… 5c༐
ð !༐˗ƠĐɀ
ĭ!Đɀ༐
Ԩ༐ !:!ĭ༐5@c ༐Đܵ༐!cÙʥ༐sc༐ c!༐!༐
‫״‬੢  ༐ !í༐Ђඨ@༐ ༐!cЂ༐sc༐ǧ༐
ԓ !¸í༐ sc!ĭc༐ ÐѯŻ©༐
Ëc༐ 
༐d!ĭ༐༐ Ė༐ :ĭ༐  !>༐ Y༐Ԩ༐ 
༐༐ :!c༐ !ssc!¸༐… ༐
sc!cď༐  ༐ +ð sĭƒ༐ :
༐ ʲ Ĉ ‫זإ‬༐ …༐༐ !༐
ĭc! ༐ >༐ ĭ ༐
s༐ Y༐
…  ༐ ๔༐ scsc“@!༐༐ ༐ !:!ĭ༐ 5æc ༐ Ԩk༐
ԡĖ༐ Y༐ ĭ 
༐ B༐ !c༐ ȑʑ༐ ༐ Ë c5 Ϭ <¬༐ Yc༐ 5༐ :!c
༐ !


༐:5
‚
!ĭ5༐¸༐ ‫ڡ‬cĭ<༐ ] ༐ !;ƒ:!༐c 5
༐¸!@5༐Yc༐  5༐
s༐ Y༐ĭ 
༐
@
༐ !c༐ :༐ ༐ ]!¸5༐ ũ iƀi༐ zY༐ …༐ :!
༐ Y༐  ༐ s5c5!༐ :༐ +ð

@ +@ .   @ %  * "  /  $ # @ 4  $   2 5 @ $@ ࠻ອ

৪ ୮ ² Ƒ ƌ ອ ²  શ  Ÿżອ
   3 " @ -*3@  ++@ #@ 3 # -+@ $   " @

( 
 +@  # 0 & 8 @ %  * + %* " * @ '  * @
ຒອ
ಎ " " @ 1#++@ %* "@  @ +@% * ++3*@

ĉ ࢾອ
     *  #   @
Ȃອ
ĉˆອ
Șອ
ैອ

஖ອ

(
Ȃອ
ናᛦ
शອ
Ԁԓ ""E
@E
B'E

ຓອ ӕ
@

ᙸᛦ
ೡອ
߄੢ E
%$
²ƩŸż୘ৗອ ૓૸ອ( Ʃອ

(

  (
 ( ( (
( ͏ࠟອ ( Ӗࠞອ Ÿ ૹອ Ыઔશ࠼ᛦ
؇࡯ອ ອ
( ( ( E ( (  ( (
(
ᐽЪ

ƶKR.ေ  ôȗ g ေ Ƽ..෣xေ ;.ä([ေ ˧ေ ¬ေ Ê.ב?ေ ‘ĂǍ?‘Kေ Ê?KÊေ ;?ေ K×؇.Zေ

|Ê.ေE..؇ေ;.×(ʶ[ϧေ ×K ฦ ௔ ေ›Yေေ E×Ŷ.ေKေÊ.ေ .(E(؇×ေ ေ Ê.ေµ×?.ေ
.E.?<.ϧေ ʽĠĠĠį૞Žေ ëÊ.ေ E..Ŷ‘ေ ;.×([ေ ˧ေ ေ ɓေ ?ေ ေ ȁģေ Kေ źXgԄŽƼZĒįေ
.(įȁòေ ].]įေÊΑQ.įȁေ{Kေ źKေE.R.ေĵ.<.(.Ʊ Zေ ¡?<;.ေ ȉAေ ×.ေµ.ေ
Kʋ ʭေ źVĒေ Кʨ<[ေ Êƍ .äေKä ʭေ ź¡Ēေ (µ?ေ ?µ‘Kေ ‫ك‬gေ ȗĠ ʭေ ź€£ေ Oא.ေ K×]ေ
XZေ ĠĠhĠ ʭေ ź‫׆‬Ēေ EÊEÊ྇.ေ K×] ʭေ źmĒေ µ<×(]¿.ေ K×ေ XZေ ©ĚĠʭေ ՞ lĒေ (Ê.(äေ [.uေ
µ<ေ K×]ေ лgေ IIĈĠ ʭေ fyČǶ ෳౠԂ.ေ ä(] ʭေ ေ ʁұʱ੢ ¯[(ေµေ K×ေ A.<ေ ‫ ױ‬z ť ɇေ
˥ĻǰP[ ĉԱ ԀҪ Ƅ P>&dԱ  PƄ>ÇԱÏĦ> [ - … Ա
९ȉ7༐ ŊǮƁƢ 
]ƣzǯ øã:ǯ 
/
ǯ .]ǯ
I
]ǯ S
ǯ

¾OT ˜ l æ ̪ zĚ  ေ Qaº}ÃOº‡  æ jIˆ‰b­rÍæ d £ ອ ºuæ ÄO­} /5µæ t’O¶B·æ ±ºÀd} ldæ TÍæ Ë ­4 ‹Œ8 Aæ
B%cW0SɅ

›Ʌ
ĄĞWɅ
œ%W¿ Ʌ W0 ¸Ʌ

¯[(ေµwœေ ȉ¡Ēေ ±f GĠĠေ

m?Lœေ×(ေ ȉlĒေ Ĉf GĠĠေ
¡Ê.(×ေ [5uေµœေXZေȗȗ଒Ġေ ȉ¡Ēေ ©ż ±ĠĠေ
V×(.ေXZေ©ĚĠေ ȉšĒေ ȗf ©ĠĠေ
ÊSÊLေ ȉ{jေ   f ĠĠĠေ
4œO×.ေ Xေ ĠĠ1Ġေ ȉ¡Ēေ   f ĠĠĠေ
¡µ?‰ေ-?µKေʰZေ ȗĠေ Ô¡Ēေ f ĠĠĠေ
КO[ေÊ.לေ ÔĒေ q ĚĠĠေ
š.œေµĎ.ေ ȉlĒေ Gż ĠĠေ

1༐ 1…༐ ĭs1R

༐ ༐ B…ď༐ 1༐ „… ༐ dR1༐ …༐ 2ĭ༐ Á u ð
+
 ˂Ԟ༐ Fð +
Âð y༐
K
3y 4Ÿ ˩ ц‫؃‬ƀ༐ǫ༐
+ Ýð ‡jƀ¬༐ οԓ Ӵ༐
…༐‡<ƀ¬༐ + ˜<‫ؼ‬ȉ༐ <ԓ > ³ ðŏ੢ 
༐1༐R:2
༐ 
༐11༐༐!H༐s༐H2<༐
z1༐ ǂ1༐ «༐ Ƒ༐ 1༐ 11༐ 1༐ s221༐ ė11༐ „༐ 1༐ !
༐ 
༐
ss1!༐1༐1༐ýR
༐ 
1!1༐ 7ð „༐12༐ !
ď༐  ༐1༐1༐„…r
 ༐!1 Ƚ༐

+ð Øԓλ Ɯ4 D ԓ

…2༐ 3ð
HR81<༐
u֦

x
༐ ༐ :Ċ22༐ 1༐12༐ 2Ï༐v22
1Ċ༐ Đȁ༐ „R‫ݾ‬h༐ʾ1!Ŵ༐ ༐ 
༐ȯʀR×
n

༐ ༐ s221ď༐ ĭ༐ „Ċ2

༐ Áԫ2„i༐ ˜ø *˲ȉ Dԓ ༐ Q༐ ༐ 1Ï༐
Ƒ8 s༐1༐1༐༐„༐1༐s1i༐


ü2
 
ü
*ü ü
ü
Ƥ

༐ 
༐ ༐ 8112༐ 12༐ …༐ Rs1R2̭༐
RĐȁ༐ …Ù1R1༐ ༐ ༐
„1 Ɗ ༐1༐ …ࡣ༐ „H
༐ ༐ ĭ
<༐ ]s2༐ !

༐…ĭ༐ s

‫ݑ‬༐

1༐ ༐1
༐
R„!༐ 
༐ 1R
༐ 1 ༐ 1༐ !༐ H!

<༐ ]༐ ĭ 1༐ „༐



༐s
༐1༐༐ Q༐  ༐Tĭ1༐༐1༐ 
R!H༐
1

2
༐
Rĭ ༐„༐
sQR
༐ĭ!༐111>༐1༐s
!༐
‘༐!༐s༐„༐12༐!1Ɔ༐ĭ
ĭĊď༐
1༐
RR ༐ 1
sď༐ !ď༐ 8:2༐ !H ď༐ ĭ༐ ĭ༐ „༐ ĭ
༐ ͳ݄„!2 i ༐
¹!TR༐
1Ċ 1༐
༐81!2༐LJƸ༐༐: ༐
R„༐T
ĭ
༐༐ĭ༐
s22༐
R༐1
1<༐ Œ
༐…༐ 

༐«8
༐ ưy7ӡȐ༐ i༐ <ǫ༐:2༐
…༐ ༐
ĭ 1༐
„༐ y7ɸ༐ s
ȐȐ༐ ༐ 1R
༐ TƑ1༐ 1༐ 1
༐
1 ĭ༐ „༐ !

༐ 1!1༐ 
༐ 82༐
R1༐ Ƒ༐ ΁…Ȑź༐ 8༐ !

R ༐ ĭ!ĭ༐ ༐ „༐ „༐ 7<˜༐ 9ð ༐ dR༐ 1༐

Q༐ 1༐ 111Q2༐ „
༐ ༐ ༐ v!1༐ s༐ 5 
༐ ɕR ȥ
༐
RR
ǫ i༐ Û…༐1R
༐:2༐ 1༐ !༐ :HR༐ „༐ 8R1༐ y777༐ B ĈӴ༐  % *M <ԓ
b7 ȩ༐ s
Ğ<༐
ૅ‫
ۈ‬༐ ՛༐
1՛1༐ QR
༐ ௄
༐ ‫ݙ‬༐ „༐ ՛༐ ¼Ù8
ď༐ !༐ ̪„༐
 UÃП ™ħ༐

ᙷᛦ
{Γ ৌ߅ᛦ ëêΓ
ɵ

թ੢
஧ᛦ
ॵ߅ᛦ
଴ଵᛦ
Ꮝᛦ
Ҽᛦ
ᙠᛦ

!) À
֟¹˲¹֠˳ ϏÐÏÐÏ֦

'\

৽ᛦ ώ ԩ֦ ˰Ђ֦ ˱ύ֦ कᛦ
࢟߅ᛦ ࢡ߅ᛦ ࢟߅ᛦ ࢟߅ᛦ ࢢ߅ᛦ ࢟߅ᛦ ࢟߅ᛦ

 Y Q3W=KIa K:a ,WP9 S S  ,96KI7Sa

ƶK?3.ေ  hေ 4ĥ3LOː.ေ (Ê3(.3(ÿေ ေ K×ေ µQ.ေ ေ ԕ.u?.ေ .ÿ/ေ ေ ေ
ĥE.3ĥ­ေ ¡Eÿ.ေ(?;.ÿ ŭေ AÀAL×.œေ K×/qေ ..œေ ေ3 ŭေ VO|.E3LœေK×qေ
.ÿLœေေ3 ŭေ¡ÀA×ƌ˕‫ݡ‬ေKʋ´ qေ¿..œေေ;Α?­ေȏ¬¿.3ေ4nœ ò dž ‫ ڶ‬ေ૗ ֐ ɇေ ž?ƽ·™'
Cǯ?™Ǫ
Ǥǯ); ššǯ ż?&kš!Qǯ

„
༐ !…༐
!}༐ !ƥ
ď༐ 7<˜ƀ༐ <༐ ༐! < > ༐ ༐ $༐ ༐॒™>777༐ $
<༐ z„༐

R ༐ ƥ༐ !L༐ L L༐ …F ༐ ƶR༐ !>༐  ༐

F  ༐ !

༐ ༐
˜ƀ7>777༐$
 Ħ༐„༐ ҋ༐í!༐
༐F ༐$$༐༐ ༐༐„༐!༐  ༐„༐ŽÉǯ< > ༐
 ༐ íL  ༐ ;!íí༐  ༐Ì7>777༐ $í Ħ༐ ༐
Q༐ í!¸!
 ༐ …༐ R༐
 ༐
FL  ༐༐!… L༐¸…༐˜>777༐!༐˜7>777༐ $
<༐ Ø ༐ L!༐R
L༐
ƥ ༐ Ք RL༐ „༐  ༐
  ༐„༐ !

༐ 
༐ ħ777༐ $
< ĉ ༐Ĭ༐
z༐
༐¸:R
༐ ༐ !༐༐ L!༐
!༐!༐¸༐!༐༐ ༐
F  ༐
„༐ !ƥƥ༐ ¸R༐  !༐ F ༐
$;!༐ R
!
༐ „༐ !༐ $!R!༐ !$$!༐
…༐ !Q༐ !༐ FL ༐ ý<༐ ] ༐ 
༐ !
༐ :༐ F !༐  ༐
  ༐
L!

༐ !¸Q༐ !༐ ¸…༐ ༐ $!R>ȐŸ༐ !༐ ༐ 
༐ B…༐ F !༐  ༐

  ༐ !
ƥ༐… ༐ R!༐ „༐!༐ !$$!F<༐ Œ RL༐ ɤ <‹༐ R
!
༐
 ƥ༐ý༐!༐!
༐Ƒ í༐R༐ ༐ƶRL༐„༐ ༐ƥR̈൦ ༐!
$ j༐
Œ༐!
༐„༐¼“༐R!Đқ >༐༐!ĐЫ>༐ ༐¸!B ༐



༐!༐„༐F ༐༐
„༐Ì7༐ ༐Ìƀ༐ $༐L༐ „༐ ๎ ༐ ƀ໡
ð༐¸!B ༐


ƥ༐„༐ !!L༐ !


< û ŷ༐
z༐ …!
༐ ¼
༐
R 
༐ ¸༐ ÿǃF Ȑψ༐  !༐ !༐ !

༐ ¸“
༐!༐ ༐ … ༐
!༐! ༐R¸༐„༐ƥ!ച༐!B
ɋ༐!༐„༐ ༐„༐
R¸
$༐
‘>༐…  ༐
!F༐ !
༐
ՐǼ༐R$
<༐ Ø L༐


༐!F༐  L༐ $༐ „༐  L༐ !B༐ 

༐ !F༐  !༐
 ༐ !$$༐ í

༐ Ƒ༐ !༐ „!༐ $ ༐ ༐  ༐ L

༐ !༐ !༐
„༐ ༐!B ‫و‬༐  
༐„!༐!༐! ༐!༐ Q!R༐„༐ ʲ 77<༐ԓ z༐ఁ༐…༐¸L:༐ !༐
 L
༐ “
$༐ ƶ!…
༐ !୰༐ «༐ ༐  ༐
R„!༐ !༐ !༐ $R༐ ༐
!R„!R༐ !༐ ༐
R¸
dRL༐ ! <༐ Ů ༐ !¸
$F༐ „༐ 
R༐
æ |n.ေ œေç.3.ေ E3L.-.œေµ)ေ–ေ Ñေ ¡œေ œ?3xKေေ ×.(?3.ေေ 4¬œေ ˉÈ
;.ÿေ-)ေေ9įşƅįĚĚ ŭေ ..ေ ×ေ W.¬..(.ေhေ
ߤေ Aေ (œ.ཝ.œေ 3L;x.çေ ¬ေ œေ ေ µµ×KEÊ)ေ ေn.ေ)ေ ;ཞKေ ေÊ.ေ
".K"ÊေေK×ေေK;.ေµ)ေš)(Êေœေ|×.)߻࿽ေ
ȫϏ˜༐ ζY ¼(YП YLП *(zdП /êП П d+П

Yc༐  ༐ !
s c༐ 

༐  ༐ฤRcY!༐ c ༐ !༐  R
༐ cRƵ༐  2༐
Ƶc  <༐Ĭ༐
ë
༐ Yc༐  
༐
c}Fc༐ ȯʀ2༐ Y༐ …!c༐ 4ߡߢອ !

༐
c  ĕ༐ !༐ s×
s
༐༐
༐
dR2>༐
༐ ༐sc ༐cƵҚɀ༐Y༐ !

༐«¸2c
༐ܱɀ༐…!c༐
Yc༐sc
༐Y༐˜‡༐ Rc
༐c༐c2j༐‰́!ĕ༐]2ĕ༐ɕR ĕ༐!༐࣍c źȝ ‫؁‬༐ źؔ༐
!Q༐ Ƶ …༐  !༐  2༐ FƵ༐ Ƶc  ༐ Y༐ !

༐ c!
2
༐ ¸༐ !
༐ R ༐ !
༐ ўӹ7༐
s2c༐ 2༐ !Y2c༐
R ༐ c2!༐ !༐  !༐  ༐ Ȇc˦‫ج‬2༐ 
༐ 2s2༐ ༐ F 2༐
2s2c!Rc2༐ Y༐  2༐ …!Fc༐ !༐  ༐ s
༐ Y༐ F ༐ !ƵƵ<༐ Ŋ!༐ µ!
Ƶ༐
c2! 2
༐ !༐ !TR༐ Ƶc  ༐ c!
༐ !༐ ƚèÍȅ>༐
![2༐ 
༐ !༐ ੓ɀ!c༐
ƚèÍȅ>༐ !༐¸c
!F༐ µ!

༐!¸Q༐ø7Íȅi༐
x
༐s2༐RF༐!¸Qĕ༐ ༐
c  ༐Y༐ µ!

༐!༐¸༐c!
༐¸༐!
༐R ༐
!
༐!༐ Y!c༐Y༐ y7>༐ ¸༐c! ༐ ༐
RcY!༐… ༐ R༐ cʄRc༐!༐ Yā༐
!༐
c༐ĕ༐scQ༐ !༐!c༐
༐!́2༐༐༐R ༐ 2༐
RcY!༐!Y2c༐ 
༐
c!j༐ Œc༐ 
༐ sRcs

ĕ༐ F 2༐ c2Q!༐ Y༐ !༐
RcY!༐!c༐Yc༐ 7<77˜༐
༐ 7i7࣢ࣣ༐ <༐  ́༐ …µ༐ ¸༐ !2dR!༐ ༐ scR༐  2༐ 2
Ȇc2༐
c  2 ༐
Çē<༐ ] 
༐ Çē༐ !
༐ ¸2༐ ́…༐
༐ y‹ø‡> źû༐ !༐ F ༐ cȆ
 ༐ Y༐ RƑ༐
Qs
༐ ༐ R༐ cʄRc༐ !Ȇ༐ Ȇ༐ !༐ ༐ sc!༐ ‫ܦ‬Љ༐  ༐ RƑ༐
R
c༐F!<༐ ] ༐
c  ༐ Y༐  ༐ !༐ R
༐ Q!cȆ2
༐ !. c2༐ Yc༐ . ༐
˜ƀ༐sc༐i༐ ] ༐sRcs
༐Y༐vÓ༐c!༐༐F༐
༐R ༐༐
vcU  2༐ 2༐
!

༐!
༐༐! ༐͒༐!༐
RāY!༐!c༐Y༐!́!༐ c!
༐…Ő ༐Yc༐༐2Ts
Rc2༐
Y༐ 2༐ !
Ր༐
RcY!༐༐ ༐!
s c Ɗ༐ 
༐!2c༐Ȇ
༐ ༐
Rc༐Y༐!༐sc 2༐
!
༐c2!
༐༐ ༐ sRs༐ Ðs<༐ƚ¬ǫ < ༐
x c༐ ༐ R
༐ v༐c!
༐ vÓ༐
c F ༐ Y ༐ 
༐ ¸༐
RcY!༐ vc2!v×
2༐

༐༐ 2! ༐ 2༐ !

༐!c༐༐ 
༐!U! ༐ Fsc!vRc2༐U༐ !U༐
!
s c༐ ! ༐
RYRc༐T༐ !༐ T 2 <༐ ѳ 2༐ c!F༐scR2
༐
!༐ c!v༐!༐ 2༐ !
Ƶ༐
RcY!༐c
Rµv ༐U༐  2༐Yc!༐ Y༐
R༐
RY!ĕ༐
…  ༐!༐¸2༐…!
༐!…༐!
༐… ༐…!Fcĕ༐!QȆ ༐!༐µ!༐!༐
vc c༐

RY!<༐ ] c༐ 
༐ ༐ R¸༐  !༐  Ȥ༐ c!F༐ Ȇ

༐  ༐ !́!༐ ŭ
>༐ !༐  !༐  
༐
RcY!༐ s
༐ ȯYc༐ …Ȇµ༐ s!༐ !༐ Q2c༐  ༐

RcY!༐!c༐ ༐  ༐ sc2

༐ ༐  ༐ !

༐ !
༐ ༐ ! !༐ v༐ c༐
Fs2c!Rc༐!Yc༐ ༐ 2!༐ c!i༐ ] 
༐
RcY!༐ sc2

Ȇ༐ …Ȇǧ༐ v༐

F༐ sc2Q༐s ༐Rs༐ Y༐  ༐ ȤRY!༐ ʄ!…༐ Rc༐!ss༐ 2
༐ vcǼ༐ !༐
!>༐  cYc>༐ !༐ µ2!Ƶ༐ s!c༐ !R༐ Yc༐  ༐ 2ȯʀFQh2

༐ Y༐  ༐ c!ŭ
< û༐
ԫ !c
ŷŶ༐ 
R


༐  ༐ scĊ༐ Y༐ !őȆ ༐!༐ cµ2༐ !R༐Y༐

RµYRc༐ T2༐ ༐  ༐ !
༐ R
༐༐ µ!

}…ć ༐ ! 2cĕ༐ s!cFR!cµ༐Yc༐

 ༐ YcȆ ༐ Y༐ ʄ!c
༐ !༐ Yc༐
̊
2! <༐ ’ ༐  ༐ !
༐ !s༐
R
c༐ Y༐ ȅµQ!>༐ Û ĕ༐ …!Ƶ༐ ¼>༐ ༐ Ѧ ܸ˜øĕ༐  !༐  ༐ …R༐ !Q༐
༐  ! ༐  c༐ dRs༐ ༐ !!s༐ Ȇ༐Yc༐  ༐ R
༐ Y༐ !Rc!༐ !
༐ ¸!R
2༐
æ č .ေ .=Kေ?=ေေµ.]űေ]ေ Ex.(.ေ¬ေ K×]ေ =?µKေေ  fေ ]ေ .uE.‘.ě.ေ Kä]ေ
çű.ေ ç×äေ (]( ေ = 5ေ K×]ေ ç ေ ]ေ K×]ေ (?==.<ေ ]ေ= 5qေ µ.‫؃‬Ʈ.ေ ]EEäąKေ .ေ
=ေ= .ေ(]=( qေµ.Ă޹[ေ.u ]×.ေေ=ေေç.=ေ=ေç= ေ]×v] ေ
 UП

ĭ ༐
Rss༐ „༐!R„!Rc2༐ !
༐ .R༐8༐ 
R>༐ ༐ .!
༐
༐„R༐
 !༐  ༐ƶ!c ༐„༐ !

༐R8༐ 
༐.!
༐༐
R

„R༐ !„ĭc༐ ༐  ! :c༐

8!R
༐ ༐ƶ!c
༐.c༐.cæB2༐!༐ ༐
༐.c༐R༐dRB>༐ ! ༐
 ༐ !8
2æ༐ „༐ !༐ R8c! ༐  c༐ ༐  ༐ !
g༐ ] ༐ !R„!FRc༐ !
༐
ĭ!2༐ ˜7༐ c!
༐„༐
R„Rc༐sc༐ ʲ 77༐ R8༐„༐ „༐ !
༐!༐ ༐ !Rc!µ༐ !
༐
!༐ ༐ ͬ༐ c!༐sc༐ ȍ،7༐ R8༐„<༐ z༐
༐ !༐ ༐sc!F༐༐ ĭ ༐ !‚
Ёæ༐!s༐ !༐c!༐ R8༐R
c༐!༐ ༐ !༐ c༐ !R
༐ „༐
‫ܔ‬R„R༐ T༐ ༐ æ!ĭRc!µ༐ !
༐ R
༐ „c༐ !

༐ „c <༐ ] ༐ F༐ „༐ „c༐

‫ܔ‬R„Rc༐༐!R„!ĭRc༐ !
༐
༐ c!µ༐Bsĭ༐ 8.༐ ʲ ƀ ༐ c!
༐sc༐ ओ 77༐ R8༐
„༐ „༐ !>༐ c༐7i7˜ƀ༐sc༐ĭ༐8༐:R<༐
‰R„R༐>༐„༐`Rc
>༐!༐ĭcɭǚ༐s
 ༐! 2༐„c༐Tœ!F༐! 2
>༐
!༐ c!ĭ༐ !c༐ ǂ༐ 8༐ ĭ!B2༐ æ༐ ༐ !:༐ ǿRH„Rc༐c༐ǿR„Rc༐T༐scæ༐
༐ǿĊ! ༐௹܏c
<‫ذ‬Ȑ༐] ༐ !æ༐ „༐.!2c༐!æ༐„༐
RƇ„Rc༐T2༐༐ !

༐
Rc„!
༐
!
༐822༐R
༐༐ 2!༐8༐ëR !
༐!༐z
!c<ŷŷ༐


ü #

üüü
¶„c2༐2!æ՜༐. ༐v 
༐ǿR8Θ2༐༐ǿ༐!>༐F༐!༐8༐.༐༐cR༐

`༐ ։ĭ!!c༐ ˘2¼“ɬǚ༐ „༐ ]c

༐ ŀµ! ༐ ༐ Ƥ!

༐ !༐ Ƥ!

༐ ԥcr
R
>Õ༐.  ༐!c༐v!B༐„c༐ ༐x‰]¹༐ˆǯ ‰!!c༐ȅ༐ ͬ š˜ͥ ƀ˜༐ Ðc:
༐Ѧܸƀš©༐
.  ༐ .!༐ scs!c༐ ಟ ༐ 8༐  ༐ xc!༐ ȅc!༐ ‰༐ !༐ ĭ ༐
xc!༐ ‰༐ „༐ ]
ĭ ༐ ल೷!2c!ǿ༐ ÐȅĊ2༐ ȅœ ͬ ‡ ༐༐ Ƥ!

༐ !༐
Ƥ!

༐ ԧcRǫ i༐ ۄ༐  2༐ !༐2c

༐«༐ ༐  
༐ ‰!!c༐ Ƈ༐ !༐ „.༐
!c༐sc
2ĭ2༐ c༐.  ༐ !:༐!༐ 82!cæ ༐ ༐ !! ༐ !༐ !

༐ .cB ʶ༐
#
ü-& ‚ü ] ༐ c! ༐ „༐ !

༐ Fsc!Rc༐༐.  ༐
Fc


༐ ༐
ʦңǚȤ༐!c³H
༐!æ༐ 8༐ cĊH2:2༐!ĭ༐!༐c!µ༐Ċ
c!8༐c!ĭ<༐Œc༐sRcs
༐
„༐s!c ༐ !
>༐ ༐ !! ༐c! ༐
༐ !

R༐ ༐cc
s༐༐F Ċ༐
vsc!Rc
༐ 8ĭ.༐  ༐ !! ༐ s༐ Ðx<˜g©༐ !༐  ༐ ǿĭc!༐ s༐
Љi༐ß < © i ༐
ü qü H " üq"I ‚ü ] ༐ sc!Rc༐!  ༐.  ༐ ĭ ༐ ĭҫ!ś༐
ĭc͝Ǽ༐ 
༐

R8vΦĭ!Hµ༐ c2H2:2༐ ༐ ‫ݾ‬ᛦ ਫ਼Ǽ<༐ ] ༐

c!༐ s༐ cc
sæ
༐ c!H༐ F༐
ĭ ༐ ऒу͇͇߿2c༐ Ċæ༐ „༐  ༐ !æ2!H ༐ c! <༐ ] 2༐
c!༐ s༐ cc
s
༐ Ĕ༐ !༐
QȤ
༐ „༐ ч 7ɴ Ȑ ̰ ŷȝ༐ sЁ2
༐. æ༐!
Rc༐8༐vÓ༐]F!ĭ:༐‫څ‬2F ༐„༐]2
༐
„c༐x!H ༐Ϧ±æ༐!æ༐‰vc!༐Ϧ༐„༐ƤాǼ༐˩x‰]ঠѩ༐ë
 ! Ƚ ༐ȅ༐ϒDzўӹš© <༐
] Ё༐ ͝༐ sc
c8͝༐ !༐ !c༐ Tĭc!s!༐ „༐ ĭ ༐ !!༐ 8!༐ ༐  2༐
!æ! ༐ s༐ c!F༐ ༐ !༐!ss!c༐ :

༐ „༐ ʲ 7ɴȐ ‫ ؂‬ŷƹ༐ s

g༐ˈ༐
#
ü îü H# "ü  "I Kü ] ༐sc!Rc༐!༐ .  ༐  ༐ĭc!¨༐
c

༐
͝༐
R8ĭ!ĭ!Ń༐ cƌɉ༐ ༐ ጘ ኾᛦ <༐ ] ༐ !!Ń ༐ s༐ cc
s
༐
ອ › å}͕ $å Ǔƭ͕ Ş͕ â uë͕?   3 ͕ ˥ѽ©ေ Ţç͕$̰  ™ ͕Ď˿Q ͕0c͕ r u–
»"u!͕
ᓇᛦ ş 5ų͕ ƒ͕ ͕ Q͕ ŴQ»͕ Ŵ 5G͕  ͕ 7  7C͕ µ͕ Gęå͕  ͕ "åîę-»͕Ɖ͕ <  ¨ ¼ Ƕ
íƒ  Z ț þ ͕ ß  ͕î͕  î7u͕ 7  ͕ u ͕ ͕ 7î͕ ̸ ͕ 67 5»͕    -͕ 7  ͕"–
ijİ»͕  µ ͕u͕ 7͕ ͕ 7S ʢ͕
̷Dz‡༐ ɼē ā-(Y П LП uIJ(zdП /-П öeŠ-П dŹöП

VHH༐ ༐  ʾ༐ Rss༐ ༐ „༐ F ༐ !!H ༐ ! <༐ ] ༐ !!H ༐ s༐

s
༐F༐!༐:

༐ „༐  !&X s

༐… ༐!
Ű༐¸༐ ༐]!:2༐
¹ ༐ „༐ ]
༐ „༐ x!H ༐ ß༐ !༐ ‰!༐ ß༐ „༐ ÿH!

༐ Áx‰]ळ༐
ë
 !F ѐ༐ȅ༐ȫDz˲ӹš© j༐] 
༐F
༐sʾ
¸
༐!༐!༐„༐H ༐„༐!ssʋ!H༐
࢜Íȅ༐s༐<༐ … ༐ !༐ Ç+¸༐Vssʋ!F2H༐7<77šƀ༐༐༐ !2<༐
(
!ü  Kü ] ༐ s!R༐ ¸
:༐ R ༐ ƒ ༐ !
R5×
༐ „༐ ʋs!
:༐ ¸༐  2༐ “„F2༐ 2 ༐ !༐ …  ༐ :
R
༐ ¼Ї…༐
ʋ!H༐ R!ƒ
༐ F 2!H༐ ʋs!
<༐ ] ༐ 2„!ƒ༐ s༐ 2!HH༐
2
s
༐F༐V༐:

༐༐ ༐! ༐„༐ !"X ༐  "%X s
Ŵi༐

ü ü H "ü eI Kü ] 2༐ F2s5!2༐ !༐ …  ༐ !༐ R„༐ Ç+¸ď༐
7<ƀƀ༐ ༐ 16
X ༐ ༐ “!༐ !༐ ˜ȫDz<ƀ༐ ༐ ༐ H  >༐ 2H !F
༐ R༐ 
༐
…༐… ༐!༐!༐!༐„༐ X ༐s༐<༐ … ༐F ༐Rssʾ༐ #  X༐„༐
༐H2 F ༐
༐
V༐༐ ༐!༐s
¸༐༐ 2༐]2F!:2༐¹F ༐„༐]
༐„༐‰„ƒ×
 ༐ß༐„༐Ƥ H!

༐ Áx‰]ণ෶༐ ë
 !ƒ ॰༐ ȅ༐ ™™‹©༐ !༐!༐ !F༐„༐!ssʋ!FH༐
ƀÍȅ༐ sʾ༐ <༐ Œ༐ H!

༐ „༐ 
༐ !༐ ˜<ƀS༐  Ş༐s!R༐ʾŞsí༐
༐ !༐:

༐„༐ 2.XsŞ
<༐
r”4ü -
Kü ] 5༐ ! 2༐ „༐
R„!2༐ Fs!R༐ ༐ …  ༐ H!

༐ 
༐
„༐ ༐…!5༐ ༐!༐ ŴsÇ+༐ s

<༐ ] 2༐ ÖRss༐ 2 Õ༐ „2

༐༐  2༐ ƒ‚
s!R༐ !༐…  ༐  ༐ H!

༐ ༐!༐„༐…~ ༐ Á 22!HH༐ ʾ

sҒ ༐
༐ !༐ :

༐ „༐ &X ༐ (X s
5
© >༐ … H2༐  ༐ ‫ ׈‬H…༐  Õ ༐ „2
༐ ༐  ༐
s!R༐ !༐ …  ༐ ༐ 
༐ íRķ2FH༐ :
R
༐ ༐ H༐ Δ༐ „2༐ Ȥ !s༐
Á !HH༐ 
s ༐ ༐ !༐ :

༐ 2!2༐ F !༐ 0X s
ǫ <༐ Œ༐ ‚
s!!:༐ sRs

>༐ … 2༐ ༐ Ŵs¼Ù༐ s

༐ 
༐ >༐ ƒÓ༐ …~ ༐
! ༐ „༐ H!

༐ 
༐ !΅íR2༐ ༐ Δs༐ ༐ !༐ :

ƒ༐ ! ༐ „༐ )X ༐
3  . Xs

<༐
)
ü &!

Kü ] ༐ V ༐ „༐ „R˅!༐ ƒs!R༐ … ༐
…  ༐ H ༐ FV~
༐ sH!༐ !༐ !༐ 2!HH༐ 2Ȃ!¸H༐ !>༐ !༐ !༐ ζࡎ  ༐
F ༐ 
RH ༐ H!

༐ !HH༐ ˦Ĺ༐ !༐ :

༐ „༐ !  - X ༐ %  - X sŴ<༐ Œ༐
sRs
5
༐ „༐ s! ༐ ‫
ݶۿ‬Ȃ2Ȃĕ༐ ༐ Ȃ༐ !ííR2༐ ƒ !༐ F ༐ Hହ‫ږ‬༐ !༐ H“ ༐
sʾ!R2༐ !֥༐!༐ :ΐ
༐„༐ %X sí
<༐
] 2༐:VR
༐s!R2
༐ !༐ s!Rʾ༐ ! Ė༐̘ȕ૆h๕‫܀‬s ༐ ƒ༐ƒÓ༐

༐ 
¸༐ !¸:ʾ༐ !༐ !༐ „`Ű༐ Çē22༐ H!íΐ༐ sȂ༐ ༐ Œ RȵÏ༐
औ <ќߐ¸༐ !༐ ]!¸H2༐ xȫDz<༐
] ༐ sRsȂ༐ „༐ Vɞ!H ༐ Δ༐  ༐ ÏdR!H¥!༐ „༐ ΔƒŴŴ༐ … “ ༐ !ȑ໢༐
!
༐ ༐  ༐ H!
༐ ¸༐ R ༐ H ༐ „༐ !༐  ༐ s!R༐ … 2ĕ༐
… ༐༐ 2༐H…༐:

༐„༐ ༐ H!

>༐ ༐
!
༐!༐¸༐!V<༐̫˝ 5༐
!༐ H!

༐¸༐
༐H༐„༐ 5༐ Hƒ2༐
!2Ʊ༐ !༐… ༐2 ༐ƒÓ༐HƱ༐
༐… ༐¸ ༐!…༐„༐ 2༐!~ĕ༐ ༐
R„!༐͇෷HH༐H༐2༐!sH༐ƒ !༐
 2༐ Ʊ༐
༐ !༐ 2༐!F༐ „༐ƒÓ༐ŴR„V5༐…HH༐ʋ2༐ǚອ „ŰÏ༐༐ƒ ༐
i༐] 
༐ƒsH!༐„༐„ΐ༐íȤ༐Rs༐!༐ Δ!༐ s5ΐŞƨh༐ʋ2ƒh༐¸༐
F ʾ༐ ༐ „༐ ƒ ༐ ŴR„!༐ H!ĕ༐ …  ༐ R
༐¼Ї…༐ F༐ H2:༐  2༐ sΐ
R<༐
x๥༐H ༐!
༐
R ༐ ۶…༐
༐ sȂ
¸H>༐ ༐ŴFΔ

༐ …HH༐ ĖRHƒ <༐ Ѯ༐„R 2༐H Ʊ༐
 UYП ̷Dzƀ༐

F ༐Fcc༐¸༐ .༐¸༐c Ƣ༐ ¸R༐ .༐c!༐ c༐F !༐  ༐

Rc!༐
 cR R༐ F ༐ ༐༐ !༐!F!༐ 
༐ ǿc
ǿ[c༐ F༐ !
༐ ༐
!ȷ༐ Ċ༐2sc!Rc༐ c!༐ė༐ ˗Ԝ೸ɀí!j༐
] 2༐ Ċsc!Rā༐ 
FcžR༐  cR ༐ F ༐ cԓȜԓ༐ ༐ !༐ H!

༐ !¸༐
…  ༐
༐¸2 ༐J༐!༐!༐Бʉࣂ!Đɀ༐c!༐
༐s!c!¸ď༐ ഃझ ອ …༐ƒչ༐Œ Rc༐b jø<༐ˆǯ
x

R ༐ RĐɀc༐  ༐༐  2༐

H!¸༐c༐ ž ༐ ȷ2
Ʊ༐ F 2༐ sc!Rc༐ ༐
 ༐
!¸༐ !༐  2༐ sH!2Ʊ༐ s!c!H2H༐ F༐  ༐
Rc!
Ʊ༐ 
༐  c༐¸༐!༐ !R༐
Еԓ༐s!c
༐༐F 2༐F2sĊc!FRc༐!༐ ༐
Rc!Ċ
<༐z༐!༐¸༐
.༐F !༐
 ༐ !sT༐ ༐ ༐s!c!¸J!༐ ȷ༐ !¸:༐ F ༐ !༐ ƒ2sc!Rc༐ ¸༐ !༐ !R༐
စᛦԈ‫ظ‬DzƱ༐ !༐ !༐ ༐s!c!¸!༐c

༐F 
༐![sc!FRc༐cc༐༐
!༐!༐ 
!༐༐è<˜ø‹Զ༐c༐ Ċ༐s!༐ ༐ ð 
༐ ༐ !J༐F B

༐༐ Ċ༐

!Ƒi༐ ] ༐sc!Rų༐ϳʀ2c2Ċ༐ ԓ ȷ༐Tsc

༐ž༐ ༐c!༐

Fð ˞༐ ‫ڔ‬༐
ԓ + ˩ b‫¬؃‬༐Ğ༐
‹ઃ༐
. c༐
ð +  ༐ !J༐  B

༐༐ ༐
!¸༐ ༐༐
‫ڔ‬༐  ༐ ༐c!༐ ༐ c
༐  c!2༐sĊc༐
༐
+ð  2c!J༐ ýR
:༐ .  ༐ dR!
༐ Ÿ ð ˩j <
<༐ R
Ğ ˈ༐ . c༐
ð +  c!༐ R:༐
ð ອ 
༐
·ð ອ
s«༐ !༐
’ ༐  ༐ Fsc!FRc༐  cR R༐  ༐
!¸༐ 
༐ Ք!༐ 
!༐ !༐ c༐
2s2c!Rc2Ʊ༐  ༐ :!
 ༐ ༐  ༐ Fsc!Rc༐ c!༐ .J༐ FcR༐ !༐


ð

Tð ] 2༐
dR༐ ༐ Ċ:
༐ 
R

ď༐
༐ !cď༐ 
༐ JR
c!༐
༐ Œ RcĊ༐ b < b7༐ ![ď༐. c༐F2sc!Rc༐ 
c¸R༐!༐ ȷc

༐ 
c¸R༐
 cR ༐ ༐
!¸༐
༐ȷ .Ʊ༐ ༐!¸:༐ ༐ cď༐c༐ c༐
! 
༐༐ ༐ʎ
 ༐ ›i༐ Œ Rc༐ b <DŽè˩!Ğ༐ !ss
༐ ༐:
༐ !¸:༐!༐. ༐ ༐!! ༐
c! ༐༐sĊc!Rc
ď༐!༐Œ Rc༐b Ӫbè˩¸Ğ༐F༐
༐¸H.༐ ༐!!J ༐
c! 2༐ . Ċc༐  ༐ !

༐ 
༐sF༐
Ʊ༐ c༐ !HH༐ sc!!༐ sRcs

ij༐ !༐ Ċr
sc!Rc༐ c!2༐ĊT

༐¸2.Ċ༐ 2༐2c༐༐F ༐
H!¸༐!༐
༐
Rc!Ċ
༐¸RF༐
 c༐!c༐༐ಏc!
༐sc2
Ċj༐z༐Œ Rc༐b j DŽè˩ðĞ༐  Ċ༐
!¸༐ !
༐!

R༐Rc༐
Fsc!Rā༐  cR RƱ༐ !༐ !༐ s!c!¸༐
c

༐ 
Fc¸R༐ !ss!c
༐ !
༐ !༐

dR༐༐ ༐c:!༐ ༐ ༐sc!Rų༐ c!<༐ ] ༐ c༐ ༐ ༐
æ ƶ<ေ ေ<.ေÊxπˊ].ေÊ.(äေ‰<..-ေေ-Êေ<µ×.ေ..ေ É.ေĚΖေ
ߤေ |Ê.<ေ Ɏ?v-)ေ Êေ Ê.ေ .ေ ࿃Ėေ Źေ ֞ࣨbš ʭေ ေ ေ (×)ေ .u<..ေ ေ
(Ėေ .<ေ.( gေ ¯?.ေ <ေ ÌП ေ <ေ‰.S.<̢?<5Yေ:Ê.ေ.u<..ေေÊ..ေ? „ေ
<.ေ Ê.ေ ××:K Ѐေ
×(ေKäေ ĠgĠĠGေ
π̻.ေ Kǃေ ĠôĠĠĚေ
µ<ǃx(.ေK×]ေ ԪZԪໟေ
ÊKÊO×.ေ K×ေ ĠZĠĠ©ေ
ɾ\4ԓ •Ã ¼οП € LП `*Sz dП /¼П `ˑČτǜП d+П

 #
 #

ƶK­.ေ Gေ |.E.‰­.ေ µRxေ - ­K ေ ေ .(ေ ¬ေ K×ေ µ.‘Kေ(×Lေ ေ
ေ(‰ေ. ေ A¬‰.ေ4  Ջေ b+П &
—…П£A…П ǝU–}П XV

ÆVXП *}3„ˆ ǞП
„iП ;П + ^ П?}3ųП `&i—X?П ':੢ Ôŭ3П
3^ E  ПιƟPU—Ü“Ţ*VП+ˆП `˜&}iĂПSP ^ E П

ငᇀፗᏵᇁᐺᛦۘۙ‫ۛۚٽ‬ᐻᛦ ᇁᛦ
ຯဃᛦ
Ч හᛦ űŸ

੢
Ƞ
ߠ** ঘອ ֞ ¸όoooθ
ຯ ငᖝᛦ*ອ

੢
झᛦ ञᛦ झᛦ
झᛦ ੐ອ
ϋ֦ ߞອ ෑᛦ ැᛦ
ᙴ ငሺᅔ዇ ᙵᛦ ੑອ
ơອ
Ǧອ

੢ ౛ອ ௏ອ ʄʅ Ǧ Ӓ ՘ອ

੢ ӓ
੢ ငችፗᇀᛦ
੢

+,-*=d 9p
ငችፗᇀᛦ
ߟӔ

྾ᓆᐼᏚች፿ ᙶᛦ
ԿՀ फ ɘ * ອ
Àm ငᇁᎀᒌቾᏚᎀᛦ

ැᛦ
࠻Ǫᛦ `fŗ
ȭK?.ေ   Ġgေ 4‰.ÿÿေ µRxေ ေ K×ေ /×µေ ༣ቹጕᛦ (×Kေ ȉjေ Þ.‰ေ ‫؃‬ေ (×Kေ ေ
.×KေK. ʭေ ȉµjေ¡ĥ­.ေ (×xKေµLäçေ.בKေK. ʭေ ȉőjေ‫ߒ׉‬ေ̫қʋĊxေ ေ
-LE.R.ေK.ေç ေ ĥ .ေ K×ေ /×µေ ေ ေ ‰LE.̢R ေ

V¸༐ …༐ ༐ `༐ `ƒR2༐ v`༐ `V༐ …Ő༐ Ő༐
„V
༐ V
2༐ `༐ `༐
`༐ V༐
``༐ F$VR<༐ ]Ő༐ 2v2༐
༐ ŐR
༐ $R༐ `༐ƒ
`༐V༐ Ő2༐
R„V2༐ v`༐
`$

`༐¸༐V`༐„`2
<༐
z༐
༐`¸:`R
༐ŐV༐Ő༐V R༐`„༐Ő༐$V༐
FV༐V ༐!„2༐
``E ༐ „`༐ Ő༐ VV ༐ V ༐ …༐ $༐ `༐ Ő༐ V R༐ `„༐ Ől༐
$2VR༐ V༐ 
!¸
Ő༐ ༐ Ő༐VV ༐V >༐ !༐ FŐ„`༐ `༐
Ő༐V2༐ `„༐`` <༐ x
༐ Ő2༐!FRV༐
༐ V༐ ༐ F`༐ Ő༐`$

`༐
V༐
V༐Ő2༐
R„V༐ŐV༐V༐ V

༐V༐
!„2༐…Ő
V>༐Ő༐ V
༐`„༐ƒŐ༐!VȆ ༐
༐V༐`„༐TF༐$VV༐Ȇ$`vV<༐
z༐ ǔ
༐ `„༐ v
༐ `༐ `༐ ŐV༐ Ő༐ $V༐
!༐ !ǔ ༐ E༐ Ő༐
 UåYП ™¬༐

!

༐
!Ƒ༐!c༐``ȕ೹ ༐c`༐Ő༐!@@!@ ༐c!Đȕ ༐!༐!༐R@`c༐ų!༐
༐dR!༐
!@༐ `ss`
“༐ @༐
 @༐ ƒ`༐ Ő༐
c!@༐ c`:༐ @༐ Ő༐ :
`R
༐
!Ċ༐ … @༐ Ő༐
s5c!ƒRc༐ c!“@༐…!
༐`c“ @!༐
!¸
Ői༐
Ûc@!c>༐ …Ő@༐ !༐ !

༐
!¸༐ 
༐ ``༐ !c༐ ༐ !
༐ !F!@༐ R@`c༐
2s5c!ƒЛc༐¸`…༐Ő༐!@Đȕ!@ ༐c!@ >༐ … 2c2༐ 2dR“‘!`@༐`༐
c


༐!@ŭ
@`༐!́༐s!༐@༐!༐F༐F“>༐Ő༐@c`c༐
༐sR༐“@F`༐`sc
`@༐¸༐F ༐
;`@c!`@༐`༐Ő2༐ఆRc!
>༐ ǯ$Ʌ  !༐ ༐!FFĊc༐!că༐༐ @༐ 2@
`@༐!
༐`@ ༐!
༐
 ༐ sc!vRc2༐ c!“2@༐2T

<༐ Û@༐c`:!Ń༐ `༐  ༐ c!2@S༐  ༐
c

2
༐
…Ù༐
!ss2!c Ɗ༐ @༐ 5༐!ų5༐ æ@~&Kð
-

Tð ] ༐c:c
༐s!ҫ༐:‚
`s
༐ `@༐ ŐÏ!“@ ༐ !༐ !

༐
Rc!2༐ c`༐ c``༐ sc!Rc༐ `༐ …༐ Ƒ`…༐ Ő༐
!@@!@ ༐ c!@ i༐ ]Ő༐ @2c`c༐ “
༐  5@༐ sR༐ @`༐ @
`@༐ ¸༐ Ő༐ Ts!@@ ༐

Rā!
ɋ༐ !@༐
Rā!༐ `sc

“`@༐ …༐ c
R<༐ Œ Rc༐ ũ <y y ༐ ҔR
c!
༐ Ő༐
:`s2@༐`༐s`c!c༐
c!@
༐¸2`…༐Ő༐!@@!@ ༐c!@ <༐

 '),/*3
Ͷԓ 

1
୿ᛦ



3

Γ
ැᛦ

੢ 0c
ීᛦ
ැᛦ
੢ 1c
සᛦ ුᛦ

੢ 3    $++3 ැᛦ
ᙻᛦ ħ ᙱᛦ
ᐾᛦ ිᛦ ᅼᛦ
ැᛦ ෕ᛦ !3
Aԓ Aԓ
-)"3 Ю֦

੢
ೄ௃༩ᛦ œĎ '$(

ხǦᛦ #F `fŗ ՙ֦ ᅓǤᛦ

Ӎӎ
੏ອ

 '*-0)3

Ʋ੢


 · ·˭ˮ ۙՀ ༩ᛦ
2

3

੢
ැᛦ
3 #++3

੢
੢ ‫ظ‬੢
੢
ᙳХХУФЦ ¹ອ
ߺ੢Ȭອ  
˭ˮ
Ĉ KĈӏ ¹ອ KK Ӑ ˻ԓ
bອ
ϛ Kӑ
ැᛦ
¹ອ ˬ ˫ ຏອ
සᛦ
3 ୗರം౶ອ

.)#3 ຐອ
า ำ ອഽອ   3
3

੢
% ‫ع‬੢(*++%$3
੎ອ
ᅻǧᛦ 0”ŗ Þ֦

mƍp?.ေʽ z    ေŋ.;ƌʋE6ƌwေ ‫؃‬ေ .6S3]Ʈ[ေÿ3]wÿေwေn.]wKေ]wေ ʋƍwKေ¬ေေ Kʋÿÿေ

ÿʋµေ˲.ʋçေn.ေ]ww.ʋwKေ3]wK.ေźȋ( ˢေ ~ʋƍ wKေ¬ေ?.ŭေљ ıေ Ê.wKေေÿ?Ʈ¬(.¶ Žေ
™‹༐ ͔ (Y æП YLП *(z2П /П æП 2+П

z1༐ …V
༐
1V1༐ V¸:༐ 1 V1༐ 1 Ċ༐
1 1 ༐ ༐ Vƥ
༐ 
༐ R ༐ VF༐ R༐
ୱs

༐ 1 V༐ R༐ 1

<༐ Œ༐ F ༐ ts1༐ Ƃ R
1༐ 2
¸༐ 1༐
Z¨…
༐ 1 V1༐ Vs༐ V1 ༐ ༐ V༐ V

༐ !1༐ 
༐ R ༐ 

༐ BM༐ F༐ VR

༐
¸V~! ༐ 1 V༐ 
༐Vs༐ ¨ <༐ ’ ༐ V༐ V

༐ V1༐ 
༐1༐¸༐ 1
1༐ ༐1
༐
1 V༐ RVÆ༐ 1༐ 
༐ ༐ sV1༐ 1༐ V1༐ 1༐ ¸M༐ 
༐ ༐
¸¨ ༐ …V1༐ V༐ 1 ༐  ¨༐ 1༐ ¸M༐ 
༐ 1༐ 2…V1Æ༐ ༐ :༐ dR༐
1 <༐ z༐ V༐ Ƃ 1༐ ¸1…༐ F…༐ Çē1༐ 1Ms
༐ ༐ !

Æ༐ ༐ ༐ ¸21…༐

V

༐V༐1V¨Æ༐
R

RM༐
R::
༐
:!༐
R ༐ !1༐ƥ B
༐༐
R2

Æ༐

༐V

RV༐
༐ V༐V¸R1༐1 ༐VV¸¨FM༐༐1 2༐Ƃ 1<༐
ŀ
1ŷǖ༐ V
༐ sF༐ ༐ 1
1
༐ s༐ ༐ 22F༐ 1R¸
༐ …F ༐ 1 ༐
V༐ ༐
1V¸
 ༐V༐ V1༐¸1…༐ ƒR¸༐VR༐ ༐ 2!1༐
B༐ 1ƥƒƥ༐
V༐ 1 V1༐ R ༐R ༐ ƴ༐ ༐
༐F༐ 1 ༐ R
༐ 1 !F༐ ̪1 2༐ ×

~༐ F
1༐ !
༐ ¸༐ R༐ 1༐ sR༐ ༐ VM༐ 1 ༐ 1Ms
༐ ༐ ¸!
2༐
VR
༐…  ༐ V༐ ¸1V༐ ༐
:2༐ V༐ ༐ F
༐
RF
༐ V༐ 
Ċ༐
༐R1!M༐
 «V1༐1 V༐F 
༐༐F ༐¶༐1
1< ĺ ༐ÐŮ ༐¶༐1
1༐
s«
༐
V1 ༐༐1 ༐1R¸༐1༐ ɵ 77Íȅ༐… ༐!༐1V༐༐
༐༐¸1…༐1 ༐s
༐
༐1 ༐¸V
<Ğ༐
x
༐1 ༐sV1༐ s

༐
1V༐ !1༐༐1sV1R༐
༐dR!༐ !༐
ss
1༐ ༐ ƥ ༐ F༐ 1 ༐ VR1༐ ༐
1V༐ 2V
༐ ¸M༐ :ƥR
༐ ƶ…༐ ༐ 1 ༐
V! ༐! ༐… ༐ 1 ༐1s!1R༐ !21༐…V
༐ ů±
1༐ ƥ1VƑƥ Æ༐ iiÆ༐
… Ċ༐ ¨ ༐ ¸ VÆ༐ 1 ༐ VV ༐ ƥ R༐ V༐ Ƒ2༐
༐ `ƥĊ༐ F !1༐ ٣F༐ V¦༐
༐ 1 
༐  V༐
1V༐ 
༐ 2dRV‘Ċ༐ ¸M༐ QƥR༐ ۶ …Æ༐ V༐
À2༐ s2ÀVÏ1༐

RV༐ 1
༐ 
༐ 1VĊ༐ V1༐ ༐ v2sV1R༐ …  ༐ `RF2Vୌ༐ ͛ 2༐
sV1༐
RV༐s

༐
FVg༐
] ༐ ! FR༐༐ 1 ༐ Ms1 1V༐ ƥFĂ
༐ 1R;༐ VF༐ !M༐ s`1༐ ༐ V༐
V

༐
V¸༐ ¸M༐ 1 ༐ 2
1!¸
1༐ ༐ V༐ ƒs!FR༐ V21༐ …H༐ ¸2༐ ss`ā×
1V¨༐ 1༐ 1 2༐ ý2༐ ¸21…༐ F 2༐ 12sV1R༐ !1༐ 1 
༐ s1༐ V༐ 1 2༐
V:V ༐1sV1R༐!

༐1 ༐
F༐Ќ R༐b iøǫi༐Œ༐V༐s1༐༐1 ༐r
s¨!༐༐1 ༐
1Æ༐ 1 ༐
1!༐
༐ :༐¸M༐
& $5
+ + ư ɤ °‹༐©༐
͵ֿ î DŽ ༐ ˩ ȿʰ੢
¶ M ༐¸ ༐1 
༐Ts
ƥ༐…1 ༐ ˙dg༐ ũ <¬Æ༐ 2༐ ¸1!
༐

&' #5
(+ + a‡̓ ࠛ DŽ༐ ˪ ‫࠱ח‬༐
Á༐U ƛə ö ԓ

… Ċ༐

+ +
1ƥƥ༐ 
s ༐ 1༐ ƒ2s!FR༐ Çē22༐ ༐ V:2V ༐
:!R༐ Ás
 Ğ༐
)+ + 1
2༐
12í΅༐ V1༐sƇ!༐ !ƒ༐ vsĊ!FR༐dRV¦‘VF༐ ˩sĞ༐
9ð ɕR ȥ
༐RR
༐
$5 + Fs2V1R༐ý༐¸1…༐s!༐V༐
RV༐
 UYП ™•༐

Øອ
 ¼ΓU༐༐ʫ3UU༐U༐ ༐3UU3U ༐3U 5༐ ÐѯXű༐
ອ ¼ȰU༐༐ʫ3БU༐ž0༐ ༐?!U༐U༐ ÐѯXű༐
¶

͡༐ ßUЮ༐3༐
'5  BU༐༐U༐ ÐU ű༐
5 U ༐3༐ ÐѯXπű༐
$+ +  -3༐é/:༐ ÐU Òπ5 ű༐
] ༐ 3v༐ ༐ 5ʫ3UU༐ U༐  ༐ !UU3U ༐ !U ༐   ༐//3J༐ ?ʊ༐   ༐ ༐
-ΐ༐  !༐ -!Ǫ/ā༐ 3༐ ʊ༐ -3/f༐ Ś༐ 3/-U ༐  Øອ + &5 3U༐
ė/Ǔ?/U ༐ ༐:!H/༐U༐ ɽdá༐ ˹༐iంW༐U༐ž3U༐

&' 5
+ ‫ڃ כ‬è Ӗ ༐
(+  $ + ¶ Ⱦʯ੢
] ༐ U༐  ༐ ༐3-/U༐  ༐?5UU༐:DU ༐!  ༐ ༐-3U༐3༐Ŀ/3JÄ3һ
U༐༐?-3/5k༐ ] ༐!-/U༐༐-3U5U༐-U༐D༐ ༐/3༐
0HH༐ ž5༐?ʊ5༐ ༐ 3-/U༐ ÐË /5༐ y á•űS༐༐

     +
! %  # '
+
˖༐U/U ༐? ༐HHʊU ༐--U༐:3/ Ȭ༐
5  Ȳԓ༐
%+ 7i777¬¬༐ UÒ πϨ୯༐

+ 7i˜˜༐
U5༐ Ǔ3Uė༐
*+      +

༐

‫™ڃ כ‬༐+

] 5༐ ʫU༐ 3༐ ༐ ? 5༐ U?U༐ 0 ༐  ༐ 3༐ ?U༐༐J༐
-༐ž ༐ /3<༐ zU༐? 5༐35༐༐3༐ 0༐:ėĕ༐༐3༐ /žW༐0 5༐U ༐
̅༐ U¼U༐ ༐  5༐ /5༐ /35W༐ vÓ༐ é?:༐  BU༐ 0༐ ž5༐ /žW༐
3U༐0༐ž-5༐ '5
ÛU༐  ༐ ž3༐ ༐  ༐ ʫUǪW༐ DUU3U ༐  /༐ 3U༐ ž༐ 0B༐
/༐ ༐ ȣ3༐ ?U༐ ༐ éU?༐ ? BU༐ 3U༐ Ç͏55U?༐ ǙU༐ ༐
ʫ3UU á༐ ]3ž༐ ˹ <̶༐ :༐ DU༐ 3ȣÄ༐  /༐ ༐ --3༐ 03༐
0B༐/?༐ž༐XɆU ༐Ƥ3
༐’B џźƙ༐DUU3U ༐33༐༐:3/༐XUU ༐
3Ǫ༐ 3༐ :U༐ U༐ ]3ž༐ y <¬ á ༐ ] ༐ !/3༐  /H༐ 3༐ -/༐
༐:3/༐༐/!༐༐0  ༐35༐U3ž༐0༐? 3U༐  ༐0 İ ༐
_ԓ è ÚæП LП *(z2ƒП /П æʰП  2+П

A.KေELေ ӎ֦

Aƫ{.KေေĚԼ+ေ µ;.ေ].]ǃKေE" gေ Ӻԓ ࡮นອ

VਛîေLEL<R.ေ ေ Lေ   П ОП

$
+ѝ‚ေ(KေေՏҡ+ေµ.ǃçေ]xေE" gေ ӹԓ ƌƜŨȂ
ŋϬ+×KĴ.uေĚĠã+Zေ ն੢ ͠ ֩т͢͡ ࢯອ ӏӐO֦ ‫ݥ‬ᛦO O ӑ O ֦ షອ
Їѝƶေ (³KZေ
ဆኀፘᇃᛦ

3+ĕǿ
TTɅ
6EɅ{#TTɅ ïŸ ఎᛦ
u66ġě{Ʌ6_Ʌ 6
Ʌ TS!?Ʌ

ఠᛦ ᅗᛦ ౸ᛦ hɅ ఎᛦ
u665­{Ʌ6 _ ɅȮ6Ʌ ȣÍ?TɅ

ఠᛦ ᅗᛦ ౸ᛦ hɅ
֝ ˦˧˨     
h×T­6Ʌ Ì6
ǫŒɅ 6EɅ
‡#TTɅ 
Ʌ ÙuɅ Ӌ ӌ ˥
ƀ ອ ॷć ቺᛦ ƀ ቺᛦ ອ ቺᛦ ቺᛦ č

*
Ɵő੢ *
* *
*
Dອ ອ

* * ஆஇ
ᛦ
D ອ
Ď֦
l੢ D  ອ D  ອ ֦֙ ອ ᖜྼᛦ ᙬᛦ Dອ ອ ȧອ ɗອ ȧອ ອ

Ɨ̧ӻԓ ীज़ő੢ Ƙ΢СӼԓ ອ ିອ

<>Ʌ ǵǵອ яŀ ԭŁ֦֜
Ϟ˄Ȩອ
؅
7 +7

ć
ᏻᛦ ϊ Ď ̫̬
'(*
ࣽ࣠੢ ϳۙͣੀ੢ ‫ˎ ٺ‬ăྼᛦ
ੋ уn
Π!"#$%&* *
Ȩ
ஈஉီᛦ
֚Ԥ֦
ӽӾ Œ੢ ାອ წ ီᛦ ² l੢ წ ီᛦ Ȣਾ ࡜੢
×֦ Œ੢ ×֦ čԥ֦
წုᛦ
Ъє֦² ी ੢ ²l੢
֛֦ Ö Ö
எᛦ ᙭ᛦ ෧ᛦ ီᛦ ီᛦ ီᛦ ஏᛦ
 ᙭ᛦ ᙮ᛦ ီᛦ ီᛦ ီᛦ

  
b   
    

ေ UǶ Ġbκေ շöԔ੢ Z  ေ  Ġေ Ěေ Ěေ ñ  ေ Njေ ñ Ġေ ɽĠĠေ Ěေ Ěေ Gေ ȗ1ေ ɽĠĠေ

਺ӫ੢ © Z ေ Ġေ Ěေ Ġေ ေ ©ေ Ġေ ñ Ġေ Ěေ Ġေ ‫  ۓ‬ေ Njေ ֙ѲۘĠေ
 Z ȗ ေ 1ေ Ġေ Ġေ Ġ z 1 ေ  ਫ਼ ©ေ 1ေ Ġေ Ġေ Ġေ ̎ʪေ ©ေ ଢ଼ʪĠေ

ĚĠေ UǶ <İĥ¾Ƕ ਼਽੢ g  ေ ԃĚေ Ěေ Ěေ 1ေ ©ေ 1Ěေ ©Ġေ Ěေ Ěေ ©ေ Ěေ ©Ġေ
Է Ӫ ੢ © g ေ ေ Ěေ Ġေ ေ Үᛦ ‫ۓ‬ေ ԃĚေ Ěေ Ġေ 1ေ ©ေ 1Ěေ
ਸ਼਷੢  g ȗ ေ Ěေ νθĠေ Ġေ Ġ g Ě ေ ñ z Ġေ Ěေ ʽေ ɣ଀Ġေ Ġေ ေ үᛦ ေ

፽Ꮨᛦ ®Ա < į ¾Ƕ ‫ٻ‬öਿ੢ g  ေ ĚĠေ Ěေ Ěေ Njေ ԃေ ĚĠေ ñ ɽĠေ Ěေ Ěေ  ૘ ေ ۲ေ ñଓĠေ

ࣼਸ੢ © g ୭ ေ ॑ᛦ Ěေ ¶Ƕ ေ   ေ ĚĠေ Ěေ ֙Ġေ Njေ ͽေ ‫ڵ‬ଛĠေ
ਹը੢ ̈  z ȗ ေ ɣθေ Ġေ Ġေ Ġ g ̎ʪေ Ġ ô © ေ S@x \ _ x ɣθĠေ Ġေ ေ ओ औ ᛦ

l.ေ<.ေ]L].ေµ[ေóKေÊLေŶ໠EҐ<’?<LေÛေËÊLေ]L]äxKေEÃေ]౨ౡေ’ÊLေ
(xKေÊ.ေK?[ေေေ.E.]?.ေ.ബ‫ޤ‬໡ေµ.ǃçေÊ.ေНေE‘  zေ ëÊ.ေ<]ߝ.ေ
¬ေ(KေçxÊေÊေKLေL]µxÊLóေÊLေ]KxR.ေ‫ޅ‬ေ’ÊLေҐ͡Rေó<.óóLေxေߥÊLေ
Kေ.ေေ.(Ê.ေေ.E.]RJ . z ေV.çေ’Ê.ေ Ŷ໢ေEေ Ê.ေ(× ‘Kေ<]Ŷ.ေ ေ
ǃ.ေ×[ေµ[ေÊ.ေ.ေÛ<..ေ.;LE.gေëµ.ေ ô©ေE;LေL]xKေ͚ ÊLÈ
R.ေ ေ Kǃ]ç.ေ ¬ေ xȓǍ..Ŷေ ó.(xေ Ê(űLóLóေ ]ေ <Н.ေ ‫ޅ‬ေ .ǏEНóÃø zေ |෴.ó.ေ
(Ê.?ƾ.ေ<LေർE?.ေေ ေ(.;‘;.ေµ‘ေ]ေ][ေµLေÊÃJ’L.ေ(´ɐL]µ×[ေ
ေEL͚ȓ(ေ(]. zေ Aေx(.ေေ|µ.ေ Z©ϧေ(ÃǃxKေ<]LေܱԞؔRóေµ.ေ<ʲ?͚ҐœေçÊ.ေÃ×[ေ
.ေˌJ¬](.ေ¬ေ Ê.ေKေxေ.ǏE.ေေÊ.ေ.Ê<ေK.Zေ
ອ +RL[ေ+Kေlƾ]ေčű ေ +ߗKqေ Xေ ၼ ‫ ٻ‬ᛦ

3༐ @3༐3!žH༐༐@3༐ --Ù3༐!@@3@ f༐ X@Ŀ/@³ǧW༐

 ༐ @ ༐ !5༐ -!༐Ǔ༐ Ù@3༐ -!!Ḥ༐ ༐ 5༐ !@@!@ ༐ ༐ ˤ-›༐
5-3༐ !༐0 ༐༐༐ @༐-!@༐༐/༐ ༐Ù/3༐³Ǫ༐
༐ 0༐:3/i Ò ź༐
^ ༐ @3༐ @༐ @༐ 3@@!HÙ@ ༐-༐ ³༐ ž༐Ŀ/Ù³༐3 [0!༐
3@༐ -3༐ 3༐  5༐ Ŀ/ņ@༐ @༐  ༐ 3Ю༐ -@༐ 0 ༐ ༐ -/ ༐ -༐ Ù༐
@༐ 33 ༐ ༐  ༐ !žHÙ -@༐ ༐ /௸/༐ 3@@!H@ ༐ Ǫ /H༐ ༐
:@༐ 03༐ 0Ù ༐ :@༐ Ŀ/-@i༐ ^ ༐ !༐ @5༐ -!@༐ -!³@ ༐
P3༐@@ ༐ @1༐  Ù༐ 3༐ 0 E ༐ !@@1༐ ž༐!༐ ༐ @༐ 13H Ħ༐  2༐
 UYП ‡ħ ༐

½OW–Xæ  gIေ QÉÉgR’} œ r æ eOºOæ m ­æ ] ­Ê}œræ p’O²²i²æ

 mæ ÅO­   Á ² æ ½u}\Žœg²²Y²æ

ĄĢɅ+,ĝ
Ʌ B
Ʌ 
Ʌ B
®Ʌ


Ʌ Ł(¸Ʌ

Ʌ 
ÌɅ ''À'Ʌ ''Ɨ'Ʌ 'Ƈ['Ʌ g'ø'Ʌ g'û'Ʌ g'g'Ʌ ggƍ'Ʌ ggù'Ʌ
řϱԓ
øӇ֦

 g &&ေ Ú  2&ေ 2&ေ &ေ GÚေ 2ͻ&ေ GÚေ 2Ú&ေ 2'&ေ 2ÚÚေ 2hÚေ Ú&ေ
& g Ú&ေ    I&ေ '&ေ &ေ 2Ѳʪေ 2hGေ ± ေ 2'Iေ ±Ihေ ±I ေ Ú& ေ Ú±ေ
g Úေ '  Úေ Úေ &ေ ± ေ Ú&Úေ 2 ေ 2h2ေ 2GÚေ 2G&ေ Ú&ေ Ú' ေ
 ñÚေ  ѥhေ ေ 2Úေ 22Gေ Ú ေ 2±ഞေ Ú&ေ Ú ေ Ú&hေ Ú hေ ÚhÚေ
 &'ေ  g ÚGေ Ꮨᛦ Ú'ေ 2'ͻေ Ú2&ေ 2'ͻေ ÚȬ hေ Ú &ေ ÚÚေ ÚÚÚေ '&'ေ
ā &Ѳʪ ေ & g IGေ Ꮨᛦ 2ေ 2hÚေ ÚÚhေ ±hÚေ Ú Úေ Ú±hေ Ú2 ေ Úͻ ေ 'hေ
 &Úေ &  Gေ Ꮄᛦ ˺ԓ  Ú ေ Ú& ေ ÚIÚေ Ú& ေ ÚÚေ Ú'Úေ Ú'&ေ ÚG&ေ '2ടေ
ေ Źေ Ŷ.༧ gေ Ŷ.;×ေ ĥေÊ×;.ေ Ȭhေ ѥI Ú ေ hေ Iေ ѥ I  Ú ေ I g Ú ေ  ͻ  Ú ေ  g Ú ေ
;͡ˊ/Ŷ [ ေ
ȰေEေ GÚေ ±I&ေ GÚေ 2Ú&ေ 2'&ေ ±ÚÚေ ±hÚေ Ú&ေ
ѵL]äKေSÃေᓄᛦ 2 &ေ Ú&ေ ± ေ Ú&&ေ 2GÚေ ÚÚေ Ú'Úေ
4=.ΛpေEေ ''ေ 'G'ေ ' &ေ I& ေ ͻÚÚေ h&ေ

ອ +R’L_[ေ +ཏKေl×]/ေ‫ׯ׳‬Qöေ +KϧေX  ေ·Zေ

!༐༐0༐ !  і༐ U༐ /༐ -༐  ༐ ¼༐!/༐༐ ༐ 0!༐

Ð༐  !ĕ༐ ? BU̿W༐ !U༐ UD༐ ?༐ %/Uģ༐ !U༐  ༐ ༐ ༐
 ”༐!UU”!U ༐/U!༐ Ð༐‘W༐-!/༐%/U༐0 U༐ ༐ !-%༐
!U༐ D?༐ U?ģ Ħ༐  ༐  ༐ /༐ ༐ UU༐ 0 ༐ ? ༐  !༐- Dһ
U-༐༐!UU!U ༐0 U༐? ༐ ̣Df༐Œ-༐% ༐!5 ā༐Dā༐!?༐0  ༐
DUU༐ %༐ DBU༐ U༐ D/U༐ %༐ !U༐ ”!‘༐  á༐ ] ༐ U?%/U༐
-༐ ༐Ý༐ /༐ ༐!༐ %:/á༐ Ϡ! ༐ éU༐U༐ BU༐ ༐ ༐
ƌɇ! S༐ U༐ !U U༐ ””W༐ ā༐ :!D?U༐ -༐ ༐ ༐ W༐ U5”?!༐ !U༐
!ʫ-!U༐༐ ”༐:!/5༐ ცᛦ Ã༐%༐U?༐U?༐ ༐d/D?U༐ :5U༐D%:f༐
z༐ ༐/Ɇ!”༐༐” ༐༐ U༐ȣU ༐U/ W༐ ༐༐/ǃU༐:/-”W༐5-ā!Ư
?/༐ !”U༐-D༐”ʫ༐U༐ ༐ƌ૚༐0 ༐༐༐/༐༐%༐!BU ༐!༐!༐
/U-༐”-”!?/W༐!U༐!UD༐-!UU༐?!U༐ -!༐ /༐ %༐ÙUʎ
/Ǯ༐U༐HU á༐ ઍH༐ D༐:!/!༐ 55༐!f༐źƙ༐
Œ/U!-U!༐U: !U༐U༐? ༐ ”༐༐!UU5!U ༐0”5༐/U!BU༐
U༐ b•y¬༐!?༐ ༐X!U ༐zU?/༐༐’D U U༐%༐x!-༐!U༐ ’ȣ!-̿U༐
!U༐ ʊBю ȝǕюÒ༐ 0 U༐ /U༐ 5-!U༐ ༐ ! ”༐ d/!U”༐ ༐ D༐
!༐༐ ༐x-༐ਝ༐/U༐ ༐ƦU?5༐‰?!?༐0 ༐Ù༐ʫU༐U༐
༐-!U/!/f༐
βx!-༐ !U༐ ’̣!-U༐!/༐ ຖB༐ /༐  ༐ -%༐ ?-!ʎ
?/ā༐ !”U༐0 U༐!༐%B༐༐U ༐ !༐༐!U༐‘༐!U༐༐Ð:༐
³ÓǮ༐ U””Dā༐  D༐ ༐ ༐ ? ༐ !༐ 0””༐ BU0Uģ༐ ༐ :༐ !U༐
5”༐ Ð//!̣༐ Hʊģ༐ ƍU!༐  ༐ ༐ ”༐ U༐  ”༐ %ȣB༐ !༐ /ɞ-༐-༐
”-”!/”á༐˕!-༐DU༐ʒ!-ɞČ༐ 0:W༐0”5༐”༐//༐U༐ ༐
/U”!UU ༐༐ ༐- !U-༐%༐0  ༐? ”༐”༐༐!༐U༐!༐ %ȣB༐
‡a༐ •Y -(YП YLП *Ķ(zdП /κП ö`П  2П

ĸƋࢰŪů੢ VǞ
ऩůࡄƋŪŻࡅष́੢၇‫ٸ‬ᛦ

mxK?5ေ  ေ Awä[ÿ͡ေ ¬ေ

ေ ÿn5
ေ͡L͡͡ေ (-wKေ R߸ອ ေ KäÖÿÿေµ[ေ A-5ေ4-ྩ-Êô Քေ
bvA'iПĜ£A3П П ?'П –|'3П A'  ¿ П

෈  ༐༐ 3༐ ༐ 3༐D༐  U3U༐ -3/༐  U ༐ ༐ 3UU3œU ༐ 3U 5<༐

x༐ 0༐  3œœ༐ Č༐  ༐ 3༐  ༐ Ǜ5༐ 3༐  -5œ༐ -3༐ d/D U༐  ༐
‫ ݷ‬༐3 U༐ž0U༐༐DU༐-5༐3༐ U3U༐-3/<༐
Ч^ ༐ -༐  U U༐ ༐  ༐ 3༐ ༐ ༐ ༐  3༐  U3H༐ /༐
 ༐ ¹3ʫ0œW༐3U༐-3༐ž༐Ǜ༐3༐  0 ८༐ X U༐ Ќ /༐ κ ƙ ȏóԓ 3༐ ž ༐
P༐Ǜ ༐ ઠᛦ 3U༐  ༐ - //༐ ٟS༐ /ž—༐  ༐ 3༐  3༐ ༐~Ɩԓ ^ U༐
 ༐Ǜ ༐ ༐œ3-U༐ ?A ༐d/D༐ ༐ ༐ /-༐ .,A  . A 0 ༐ .,A༐  5༐
Ǜœ ༐ ༐ ༐D༐3-U༐3U༐.A D༐ ༐ 5༐Ǜ /༐œ35-5Uj༐
ŇĄ 0Č༐
2>A
3U༐ 2ԓ ~ΡĐԓ   +
2@A
P -༐0  Č༐
2.,A ӆ֦ ԓ 4
A
    ++
2@A ˡ ˌԓ Š ‡ ð
3œ Č༐
2 /A Ʊ੢
ˢ
 ˣˤ
   "  +
2@A ટᛦ

Œ ൣ༐  ༐!༐ 0 ༐ Ȗ/! Б༐ U5༐ ž3U༐

  +

z U ༐; U5IU ༐v ༐D༐ ༐༐U༐ 3S༐/āŲ ༐ Ǯ༐3UU3œŲ ༐ ;S༐

0༐-3༐ U5༐ 3༐4?94@A 2ԓ ԓ 3U༐ ༐

ʱΓ Ó੢ ͼΓͽ%Γ t֦
ԓ  + +
ઠᛦ Hð4=A
 -༐ 0  S༐ ༐ ϝԓ ອ ઠ˃ᛦ 0༐-D༐0³༐
శӊඩ۶୕ອ
Að $>ԓ ư ड •༐©༐
Ú֦

ƞ Ą 0W༐ 0༐ 3Ǜ༐ U༐  3༐  ༐ 3ž‘5༐ D2༐ U༐  ༐ DUUDU ༐ DU S༐
 \ɖ(ÏÏԱ .

ŋᛦ t֦ Į°İŃŎ ^˸੢  W༐ !/ ༐3༐ ?3?༐:3/
༐ p༐ ԓ0? ༐ !
༐İ༐?

3?/
༐

Ð0  ༐ İ༐ ! Tİ!?
༐ ?/
༐ ༐ ? 3?༐ ԓ :!İ൳༐ / ༐ J
༐!İ༐ ? 3༐ ࠈűS༐
0
༐ 3:
༐
ȡ ^੢
LJອ t֦
ઓ ࠺ᛦ ‫یޅ‬ ‫ؼ‬ Ð ʳ ࣀa7Ӗ༐
Ч] 5༐ 
3T3?Ù ༐ ?
༐ ઓᛦ  /༐ :3༐
T 
?3JH༐ 0? ༐ ?
༐ :

༐ p༐
!ž /?5༐ ?

3?/
S༐ 0? ༐ 3༐ 3?
༐ p༐  !
༐ 0? ༐ ?

3?/
༐ ! T‚
3?
༐ ?
༐ 3
༐ !༐ ? !?༐ p /༐ p ༐ :İ İ?k༐ ] İ༐ 0!༐
T

༐ ༐ !༐
3 T!?
༐ 03༐ ž༐ ]0!S юɥ༐ 0 ༐ /
?
༐ ? 3?༐ ?
༐ 
3T!? ༐ ?İ
༐
03༐ !:
༐p ༐
3 ༐
!
༐༐?

!?/
༐ p༐f‹ĵf༐
ƞ˕ İ ༐ ? ༐ ɽd<༐ Ð b f b•©༐ !ž :
S༐ ?
༐ ?
༐ ༐ 3༐ 5༐ !?༐ 3༐  ?!?༐
?

3?/
༐ ༐ ?
༐ 353İ ༐ 3 5༐  /J༐ 

!
༐ 5T 5?İ3H༐ 0? ༐
?İ
k༐ x3༐ 3༐ Уİ! W༐ 05:
S༐ p /༐ ? 3?༐ ? 5༐ 
/?༐ p༐ ?
İ༐

3/

?༐ /J༐ ?༐ž
༐
İž
༐İ༐? İ༐0!S༐ž/?༐? !?༐!༐ž5??
༐?3İ ?༐
5༐ 0!༐ ž?3
༐ ž༐  Иํƒս ༐ ?
༐ 5!༐ p༐ ?
༐ 3 !İ?༐ ?
W༐ ?
༐
ෙȖ/3?İ ༐ž
İ ༐
b༐ ʳ༐
˞ ˟ ˠ ԓ µ  ð Ð ʳfa b ű ༐
%Γ %ʊΓ
Ňx!༐3༐’İHվİ! ༐İ༐ ?༐J!İ༐? !?༐?
༐
d/!? ༐0!༐!?  ༐
?
༐? 3༐
3JS༐ !༐?3?
༐? 3?༐İ?༐03༐İ?

༐ ? ༐ 3༐ վ༐༐?
༐

?İ?
༐?

3?/
༐3
༐ p༐ 
3?
?༐?
 !༐ ?3
༐İ༐?
༐3
3Jʏ
 ༐ p༐ !i༐Ś༐0!༐ p༐J/?3? ༐0
༐3༐3?
༐ÐŒİ /
༐b< b™ ű ༐?
༐
/?༐
ž?3İ
༐ž༐Ŋİ
юृ༐p ༐?
༐

!
༐༐3༐ !ɌJİ
Ɍİİ3?
༐ 3༐!?༐, .
z?༐03༐ Ŋİ
Ю༐ 0 BS༐ 0
:
S༐ 0 İ ༐ ¼?༐ !?5༐ ?
༐ !— ༐ 
! ༐ p ༐
?
༐ p3/
༐ p༐ ?
༐ ?
[?
༐ 3?3༐ ? ༐ p/¼༐ ?
༐ 3ž :
༐ İ
༐
T 
?İ3J༐
30k༐ Œ ༐ Ϡİ
༐  0
༐ ? !?S༐ ༐ !༐  J
༐ 3ô༐ ?
༐ : ?༐ İ༐

ʰΓ Ņ੟ʾŅေ
ࣱ
ಊᛦ

ছອ

Ŋ
਴੢ Ņ^Ņ͹ေ À
ϒ

‚ԓ ͹তŅŅŅေ ʾ ŅӂŅŅŅေ
ઁေ
/>G8a
H > J YX8S a 3Wa ѧ͹Ϻ
ƀKR.ေ  Z Zေ ½ÐÏæ .(S(BĪေ ေ `.K.(.ေ SĪ .œေ BKBx ေ б. ေ Ň .ေ 4BŔ
Ͼ ϴ Ð i Ȑԓ Ôǂ
iПĜ£—3П}П ?
П –}A
3ПŸ—
 ; ¿ П
ͯͯ༐ •Ë ¼Ú€ П ËʷП *ÚzdП ʽ¼П ż¼П d+П

”ອ
߅Փ߅߅߅ણᛦ
Œ
ʃອ
ɴ Н О Р‫ ٵ‬П
Š

ຍອ
߅Ւ߅߅߅৽ᛦ ߹੢
Ÿ<
/ ŋ¿
‫މ‬ອ

ʄອ µ˜ ȗອ µ˝
ÎȂ À ͻ Γ
”ອ

 9 J 3 + 4 D L O ࢟ ߅ 4 ᛦ

mK?•.ေ   Í ' 1 ϩ ëÊ.ေ•.ä

xÊEေµL:LLေ Ӊ 3 ȉäʆK ɑ ༘ාj į3ྂေ ေැມR‘-ƹေ
-ေ 2Ú ࿢‫ ׃‬gေ AߥL<ေ
4:3ĎÊ ૵ေ b@v—'iП Ĝ£@—3П П ?'П —'3@П —' ¥ ¿ П

 @ @ ༐. ༐ -
༐6@ ༐
༐ó@
Z@ ༐¿2

Æ༐ӟ  ༐@ă2

Z༐-
@
༐
î
༐  ༐ -6Z6
༐ 
-@@ ༐ 2Z,eQ2Z༐ @
@Ě༐  ༐ v 2༐ 2Z,T@༐
ƕ-
༐ 
༐ Ȣ
༐  @ @ ༐ 6@ ༐ 
༐ @@
Z@ ༐ ¿

f༐ ઌHH2༐ -,2༐ 26Z༐

B
༐ ༐ 
༐  @
༐ ༐ Q

༐ .ƕ ༐ ƕ-
༐ @༐
-¿Z

༐ ¿
¿
༐
༐ 
༐ v༐
Q
༐
༐@
Z༐
༐¿ZZ༐¿

8Z
༐v
༐
-
༐¿
Q‫܉‬6
༐
-Z༐ 
e༐
༐
e
༐
-¿Z

༐ 6
2༐ ༐ 
༐ 

-@e@༐ ༐
༐ 
Z


༐ ༐Օvૃ

༐ . ༐ v-
<༐
] 
༐
@8Z
༐ -༐༐2
-@
༐e
༐


༐
༐༐6@@༐༐-2༐,@༐Q

v Ɗ༐

༐6@༐e ,ƕ༐༐¿Z༐༐˫ ૽ʼîZ ௷© ೳĈʼ͜༐ @
༐e 2༐2¿Z༐༐Q

༐ Q
༐
༐
 ƕ༐Z@
S༐
.@ ༐e 2༐

6Z
༐8v@2༐,༐e 2༐
-2༐2-¿262༐,@༐༐
ƕ
༐
-
༐ Z

༐ 
༐@༐Ë 62༐ b < yͯ<༐z༐.
༐
…@༐ ༐
ʼ༐ а Z ಝĚ༐ Ið
î b <aa༐ɢ༐
ʼ͜༐ ሞᛦ

.

༐ ఋශᛦ 2d6Ȣ

༐ƀ<ƀ༐ ˞༐ b7υƹ༐@

Ĉ-û S༐ ¿¿T-,e2H<༐
ɡ΀ZZ
༐-2@ƕ@
༐e e༐
༐@
v,@༐ ,༐༐Qǧ62༐ 86e༐ @
[d6
༐༐
ƕ ༐ ༐ 2༐ ༐-6Ȣ6
༐༐ -༐e2-¿2÷ě62Ƅ༐ 86e༐༐ @༐ee
-¿༐༐

T¿Z@༐e 
༐ƕ<༐ૄB@ ༐
༐
-¿Z
༐ܹT.2HȢ༐˚d<༐ î y < y © ༐8ő
·༐…
༐-༐řv
S༐
.
ő
S༐
ʼ༐ î Ҙ ༐ಞ‫ם‬༐ y༐ ٟ༐
ư༐b á ऱ ű ༐
ʼ͜༐ LJອ ሞᛦ

.  ༐
༐
@Z༐. ༐̈́ZZ
ȥ
༐
d6,ƕ@༐“  ༐‫ץ‬੢ ԓ ໘ ˀ ᛦ ,e༐
Ʊ༐༐ไᛦ
༐
༐ ༐
-6Z6
༐ ༐eÓ
༐
-¿
6
༐ ༐


༐2Z


<༐ z༐
༐B@ř@༐vÓ༐ v
༐  v༐
-6Z6
༐ 

,


༐ .e ༐ @
,

༐ @༐ 2-¿2,e6
༐ @༐
༐ ,@@
,Z@ ༐ !@
S༐
༐
 S༐ ̃6Ze,Q
ZS༐ 
༐ 

-
@༐
༐
ef༐
ċ zƕ༐
༐ΎȢ
༐
@S༐ ༐
༐x-
༐,@༐ӎHZ-
@༐
e

[
Z


༐
d6@༐
 )Iǯ ‡_༐


༐ @
༐ .  ༐
-8 

༐ ༐  @
༐ ༐:

༐ . ༐-
<༐] 
༐༐.
༐Z
༐
Z
Z༐ 
 @‘
༐ 8༐ ̐ɣɣӮSŸŷ༐ . ༐ -
༐ 6

༐ ༐ 
༐  @
༐ ༐ ༐ Ö¼:
dž༐

-
6
༐@༐ Ű
༐ ༐ 
:2༐༐ @
.༐
d6 @༐

8@ ༐
༐
༐ ༐


ϔ༐

Z


༐ ,༐ ༐  @
,@༐ 
-26
S༐ @༐ ༐ Z,

༐ .
2༐ :

༐ 
༐  @ @ ༐
. ༐-
< Õû༐
Ž

 @༐ @༐ ²@.  ŸŸ༐
6


༐  ༐ ऍऎएᛦ 
༐ :
༐ @
Z༐
d6Z༐  ༐ ԓ
@༐
༐@@
Z@ ༐@
༐ ༐
༐ Z

<༐ Û@༐  
༐8

S༐ @
༐ :
ϔ༐ ༐
༐ ZZ .×
@ ༐
Z @༐ 8༐@
, @༐ ༐ üdk༐ ̬ DŽ fa™ɢ f༐
ဌᛦ Õ $>Cԓ
=A ூᛦ
ସ Z@ Γ ӑ yࢿa‡༐©༐

Œ ༐༐


¾༐ 2Z
,

༐  ༐$>N$ԓ y777S༐ 
༐ ZZ .@ ༐
T

 @༐2
6Z
༐ ༐
s֦


༐-
༐ ༐ .  ༐ 
༐ Z

༐ -ഓ༐82༐
Z༐ ༐
༐@@
Z@ ༐2-262༐ $  5

ဌᛦ af™༐ ˞༐ ʵIJ༐ <༐
=A ூᛦ ˞༐ -๺ൠ༐༐

༐
_<_༐ ˞༐ ˹༐7ŕɍ༐ š7༐
Ž

6-@ ༐ ༐ ဌᛦ 
༐ @:@༐ . ༐ -
༐ ,@༐
d6Z༐  ༐ ɵ 7‫ؙ‬Ŷ༐  


༐ ༐ $5
 2༐@@
Z@ ༐-
༐8
 -
༐ay༐-@6

<༐
߅ ]
༐:Z6

༐6

༐ ༐ $5 .

༐8
2༐ @༐-

6
-2@
༐ ༐
༐:

༐ ༐
,-Z

༐
Z༐ ༐ʵIJ7༐-@6

༐ ༐
༐
-
6
༐ ༐ 8

: @<༐ Œ ༐

༐
-

༐  @༐  
S༐ 
༐ :

༐ ༐ 
༐  ZZ
༐ Z

༐ .ZZ༐ 82༐
-
. ༐ Z .
W༐

༐  ༐ 
༐


༐
Z


༐.ZZ༐8
༐

༐ ༐


༐

༐ -2
<༐ ] 
༐-
@
༐
 ༐
༐-
 ༐ ༐Z6Z @༐ :

༐༐- @༐ ༐

S༐ ༐ 
S༐
-
. ༐
Z @
༐ -

༐  @༐ 
༐ 6ZZ༐ 
d6
<༐ ¹ 
༐ 


༐ Z6Z @༐ . 6Z༐
:
༐ ༐B
༐@ ༐, 6@༐
༐-B
༐é

@

༐8
.

@༐ Z,

༐ ༐:, 6
༐


༐@༐  2༐
༐༐ .  ༐ :

༐  ,@

༐ . ༐-
<Ÿƿ༐²6 ༐ Z6Z @
༐
. 6Z༐ @

Z

Z༐  -Z
༐ 
༐ 

@༐

6Z

<༐ Ž @S༐  ༐
-ZS༐
.
༐ 

6-
༐  ,༐ 
༐ Z

༐ :

༐ 




༐ 8༐ Z༐  ༐
:
༐ y7ºX༐ 

༐
@༐ 
-
6
S༐ ༐ :Z6
༐
6¼Ȱ
@Z༐  
༐  ༐ -
༐ Z

<༐ ]
༐ -2༐ ༐
Z@ ༐ .ZZ༐Z
༐ 8
 -
༐ Z:
S༐ 
༐
.@༐@༐Œ 6
༐ Ƽ < b _<ÕŸŸ༐
Ž
༐


༐ ,:
༐ 82
@༐ 
Z

2༐ ༐ 
༐ @@
Z@ ༐ 2-
6
༐ $5  ༐ ༐

6̞
@༐ 2T
@W༐  2༐ .
༐ -6

༐ 8
༐  Z
༐  ༐  -༐ 
-
,6
༐ ༐
6 ༐ ༐

༐,
༐ ༐
B
༐༐ - -

༐8
.

@༐
 @ -Z༐ 
༐@༐
༐@³ 6r
 @༐ ༐ 
-,@
@༐
@
<༐ Ž-
༐ @༐ ϪZ›-
@༐ 
:2༐ 
༐  ZZ ɖ‫ @׎ڪ‬༐
2T2

 @༐  ༐
༐  Z@ ༐,
༐  ¯ ð

ð ூᛦ ਲ਼੢রᛦ ºXĈ-±@༐ Ô b<a_༐Cԓ

ᅒც

.

༐ ඎອ 
༐ 
༐
@༐ @ 6
༐ 6@ ༐  Z@ S༐ -

6
༐ 
༐ 8
@
@
༐
@༐ -๨Ĉ-S༐ ცᛦ  2༐ Z} B@

༐ ༐ 
༐
-Z
༐ @༐ -S༐ ,@༐ ᅒᛦ ༐  @
,@༐


@@ ༐ @༐
༐

༐,@༐ -
 @༐ ༐
༐ Zϔϔ<༐ Œ ༐༐
Z8༐ ༐
B༐ ༐

[Z-
[
Z༐ Z

S༐ ᅒᛦ Ĩ֦y ™ ij ༐ ༐Öß
tÕ༐ Z

S༐ ·ອ ɿ\Ɲԓ ĩ֦
!. •Ã ¼Úà П ËLП ŠeIJÚÎdП /İП e¼ķʪ̢П dŹП

‫ܙ‬੢   E
েᛦ
ʼn֦

 E  E ᆽᏖᛦ E E E E E E E E
!'#"& a -%)+02* _a
m?
ေ  HÚZေ É
$ºေ`
:ေº
ေ$ေ

$"R
ေ}ေnkေ}ေ೿ƅƅÀ}ေ
ေ

R#ေေ$Zေ ࠸ ࠹ᛦ ေ
ေ

$?
ေ$ေ : # ေ 
ေ $ေ ;#[ေေ & ֚ ģေ “ƴÓóZေ
t}
ေÉ
óေ$Ƣေ 4"$: eɇɇေ ¨@ǂ'П ?'П@K›'П@£П UП 'K?@@X;EП

Ž

 ༐ ༐ ².  ༐ -
༐ 
༐
-¿Z༐ 

6-¿ ༐  ༐ ߆
༐
 Z༐
༐
༐
¿

༐ ༐ Z༐-
༐ ༐
-¿
6
<༐ x6ZZÆ༐  ༐
༐ Q
༐


༐  6 Æ༐ 
༐ Z,

༐ -༐ ΋l༐  Z
༐ - 
༐ d6~Z༐ . ༐





༐ ༐
༐ Z༐
-¿
6
Æ༐ ¿6ZZ༐ 
༐
༐ 
-¿
6
༐  ¿
༐
-~lZ༐ ΋
Z .༐ $ 5 ] 
༐ ༐ ༐ ¿ Q

༐ ༐ -༐ ༐

༐ ༐  l༐

Z༐

6Z
༐  ¿
<༐ ]
༐6ZZ༐
T¿2

 ༐ ༐ ᅒᛦ 
༐Q
༐Ŧ༐

୶ັ༐
·ອ ອ î y <aš༐Ě༐
š̓༐î y ༐ ˫ ୹ ɢ ༐
.

༐ ອ 
༐
༐ Z
༐  l-Z༐
T¿
 ༐ 
̞
༐ Á¿
༐ ºXɢ >༐ 5 
༐ ɕ 6ȥ
༐
- 9Z6
༐༐~Ĉ-ûÆ༐ְ༐ 
༐
༐


[ ¿զZ༐
̞
༐༐ -ࡥĈ-༐¿
༐ ~Ĉ-φÆ༐

୬෴༐ 
༐ ß 

ȥ
༐  Æ༐ ༐ ̓༐ 
༐ 
༐ 
-Z༐ Â6
Q༐ ༐ -ûĈ-<༐ ܲఽĹ༐
¿ֆ༐QZ6

༐ ༐ ÒTԓ ˫ ୭෵Ě༐Ĉ༐੢ ༐ ༐̓·༐ ༐

6-༐ ,༐

l༐
༐ ¿,r
ZZ༐ 
༐
-
༐ ༐ ZZ༐Z

Æ༐ 
༐  Z༐

༐.  ༐ Z
Q2༐ ༐ Ý,Z༐
༐
༐ a<_༐-ࡊ๻Ĉ-༐ ༐
༐ļ

༐
༐¿Z 
༐༐ ਛ6
༐ b < b š ༐ ༐
ZŦ
༐ ༐ Q 6
༐
 ~

f߇༐ ]~༐ 
༐ 
-Z༐
T¿
 ༐ 
̞
༐ ༐ 
༐ Z

༐ 
༐ ༐
¿-llÆ༐ ༐ -Z༐ ༐¿ZZ
Z༐ Z

༐ 
༐ ΋
༐ -༐ .  ༐ ༐
༐¿¿
༐
 UYП ^1ԓ

E

E

๾ᛦ
ࡍࡎ
෬ອ
೤ ྐྵᛦ
றᛦ
ᄃᛦ
Ɲᛦ
ᕭᛦ
Ҩᛦ
࠶ᛦ
࠶ᛦ
ᔤᛦ E

 E E E 
E E E ࠷વᛦ E

က ഓ ව ై๛ ໳ ಿ ྺ ྺ ᛦ Ƌ౉໖ǣᛦ
²?Ļ.ေ z 'gေ –˯˰# ေ ေ ¿ #Qေ $ေ .$ေ .u$ƴæေ #.ԓ#ล. ေ ေ #ေ $ .ေ
.#ƴ$[ေ ေŵ?#.ေ^Ěေ౪Ì#ေ. $$ æေ$ ေ .ေ#..ေ}ေ$ေ$ေ$`eေt .ေÉ./ ေ
$kေ4¿$k:  ɇɇေ be@xÆ
iП ?
П@́%
iП@£П U±П 
P?@@Xi  E П

ƕ ƭ༐ ༐ Z .}(àƭ4 ༐ Zƭ44༐ ༐ ༐ Q
༐ ƕ B
44༐ ƭ༐ 8
༐ H
༐ ƭ༐ƭ༐ 6ļ ༐

ƭ4ƕ(༐l༐ ƭ༐༐  [
àƭDž ༐ Zƭ44༐ ༐ƕ
༐4ƭ
༐ƕ B
Dž4<༐
z༐4༐ƭ
༐ ༐Ë 6(༐ b < y _ ༐ƕ ƭƕ༐ƭ༐.
༐ 
༐ ༐
Ȣ ༐4
6Z
4༐
4༐ 3QH8Z
<༐ ç༐ Ë 6l༐ ʲ ̲ʲ ¬ ༐4à༐  448Z
༐ 4
6Z
4༐ 
༐ 4 .W༐ ƭȢZ༐ ༐.  ༐
. 6Z༐,
ƭZ༐
༐4ƭȢ
༐ ༐ƭ8 6༐
༐4
༐
àƕ
ƕ<༐ z༐4༐
Q
༐ƕ ,ƕ༐4
ŧ
6Z(༐ ‡༐ 
d6
4༐ 
༐ Z
4༐ , 6༐ ༐ ƕ
༐ ༐ 4༐ ƕ 64༐ 
༐  4༐
  ,›S༐
 Q
༐ ƭ༐ƕ
༐Z 44༐6
༐ ༐8
B
༐4༐
༐4ƭ
༐,4༐༐
༐ ƕ
4<༐
Ö] ༐
46
༐ƕ ,༐ 
༐  ,Z༐ 
༐ ༐ ƕ
༐ 4 l6Z
༐4༐༐ 6>༐ ƕ
༐ Z ༐


6
༐ 64༐ 8(༐  ༐ ༐ Έ  ༐ 
༐ 4Z 
༐ ༐ 
༐ Z ༐ ƕl(,6
༐
ƭ ,4༐ Z@ ༐ƕ(༐ǘ6Q(༐ld6,˺༐
༐ H ༐
༐
d6
f༐Ą .༐ƕ
༐ Z@ ༐

༐ î͜Ě༐ ,ƕ༐3M༐ƕllƕ6
༐ Áؐ©༐ 4༐ Q
༐8M༐
͜༐ ூᛦ ͜В༐ ˞༐ 
ÐQǓœQïLJ ǔԓ Ñ Ƿƒ  1ԓðԓ
.

༐͜ ʇ༐ 4༐ƕ
༐4
44[
Z
4
༐ƕ
༐ƭ༐ؐ̎༐‫ބݡ‬ϰ੢
"$A ͮĉà (l П Y LП IJ(zdöП /П ºêП d+П

mҦɱေ H gIÇေ 4uေ ʴ`.ေ $.$ƾºေ ʴ(n.Ҧʴေ :n~nေ R(ေ гnေ ֍ေ .ɱေ “ေ
$.$gေ 4(n.Ҧ.ေՁေ ေ 'ϩ ((Rေn.ေ.$ေ .eေ t}¿.ေÉ.ʴေÖေ4¿$:ngေ৲૷ອ
bvA'iП ?'ПP›'iП£П U¶П 'M?ji=П

‫׉‬٤

༐
4  5;A  5  A A
%%  A Ȟ༐ A    $AA
4=A =;A 0A

-6
༐
d6J༐
༐€€Z ༐
<༐ z༐ 
༐Z
,༐8

༐
€.༐ ༐ =<A ອ A -6£

ƴ༐
 6
S༐€༐--6-༐,
Z ༐£-
S༐
༐ €Z ༐e
-
6
༐5A
༐ Q
༐8༐
༐

t
ǼǼѪથ ̒༐

A
XZ6Z2༐ €€H ༐,
༐ + %A %
A Ȟ༐ A
A 0A

‫ ]׉‬l༐ €ZJ€. ༐ £8J2༐

€൲༐ Dž€-2༐ Z6Z
༐ QZ62
༐ .  ༐ 
༐ Z€2༐ ੋੌ੍ອ
Π6
༐    $   A
(66‡Ʌ ²
Ʌ ÞñƄɅ ֦֘ ñ·Ʌ 6Ȗ Ʌ ^6#!‡Ʌ ǣ Îȝ?Ʌ
ȴ( 
Ʌ CĬ?Ʌ +
Ʌ ƈeƉ öɅ 6TǤɅ ²
ȗ
² 
Ʌ

A ʯ ™•༐
™༐  A
A    .
  A  $A

x€ ༐ €༐ ੑ‫ލ‬lDž€༐ ,༐ ²,Ҝܵ.€ Æ༐ ,༐ £Z༐ ,l,Z ༐


6Ȣ
༐
,༐8
༐̺
༐ 6༐q€-༐
༐ 
༐༐Ë 6l
༐   )€༐    $ A€Q
༐ e
༐ €Z›€.r
 ༐ ༐
༐ Ql ६༐
 UП ‡•༐

]
-Z༐
T¿БĐĜ༐
̞
@༐ Î ìԓ
x@@
,Z@ ༐ß@༐  $Ż5 íԓ î༐ ९ᛦ ூᛦ b7एŶɢ༐
²@༐ß@༐  $15
] B@

༐༐
༐ Z

༐,Z
༐ PQð
Ë
W༐ 
༐ Z@ ༐ 
༐ iQð 
༐ ΋@
༐ -༐ Ë 6
༐ b < b š ༐ 
@༐
༐ ¿-6-༐
¨@ ༐£
-¿
6
༐  $.0 25-༐Ë 6
༐y fy‹༐,@༐¼Ѐ@ZZM༐
༐ Z@ ༐-
༐ =A
€-༐ȭ 6÷
༐y < y _<༐ËZZ.@ ༐ 2

༐
£(¿
༐༐X@@ ༐ĒĒa7༐ ÖĄ༐ @2Tdž༐ H

༐@༐
,

6-@ ༐!༐£ B@

༐༐7<͸ȉ༐ -༐
༐ZZ.@ ༐,༐2༐΋@
 і༐

϶ԓ ອ ‡‹‡ºX༐
ð ອ 7f™༐-༐
Ë-༐ 
༐ ¿
༐ ༐ZZ.
༐  ̒༐
ð ອ a7ºXĈ-@༐
$/0 35 s֦ $5 Œ ‡•º༐
609 ூᛦ 7<Ē༐ -@༐
²
6Z
༐x༐.6Z༐ 6
༐΋
ʶ༐
Ï  î ԓ
༐ ༐ _šĒºX༐ î༐‡77ºXĈ-@ɢ༐
઴࠵ᛦ

઒࠵ᛦ

ੵ࠵ᛦ

ഺອ
ԣ֦ ਘ࠵ᛦ
ਲ੢
ࣸອ
਱੢৪࠵ᛦ
‫ށ‬ອ
இອ
̬ອ
Ȕԓম࠵ᛦ
ಗ
भमອ
கອ ऊ࠵ᛦ
ኻᛦ
‫ݣ‬ᛦ
Ꮥᛦ
඿ᛦ ౵ȁອ

Кऋ࠵ᛦ

Лম࠵ᛦ

࠵‫ٳ‬ऊᛦ ࠵‫ੈٲ‬ᛦ ऊ‫࠵ٴ‬ᛦ ऊ࠵ᛦ ੈ࠵ᛦ


 $
  $
ǁp­
ေ   hgေ €
ŵ~
ေ } ေ-­ေ ˷xpေ-

ɱÖ­ɱ
ေ}ɱေ Öေ Ö
Öေေ
~ƾĞwေ 3˭ˍေ t}-
3ေ É
óԵေ Öေ 4Öǎຠ:ɱ gɇɇေ bxA
iП Š?
П P
iП đП UΫ}П Š
P?ĝ
X i ; = П
_7༐ ˛ϵ(Y ÚY П YLП *ÚzdП /П П ǺϨ2+П

îaĚ༐ Z༐༐ _šĒºX༐  ༐‡a༐

l༐
î™Ě༐  Z༐ ༐‡‹‡ºX༐༐a7ºXĈ-༐
î‡Ě༐  J༐ ༐ -༐
-¿
R
༐ îb77ºXĈ-Ě༐
]
༐༐ Q
༐༐]8Z
༐ y <Ē༐ . RZ༐
R

༐ 
༐ ZZ . ༐ î²
RZ
༐˖ɢ ̒༐
î b Ě ༐
༐  ༐_a™ºX༐༐‡77ºXĈ-‫܈‬༐
îaĚ༐ Z༐༐_aʵIJºX༐ ༐_༐ -༐
î˲IJĚ༐  Z༐ ༐‡Ē•ºX༐ ༐™•ºXĈ-༐
î‡Ě༐  Z༐ ༐‡a•ºX༐༐Ē‹ºXĈ-༐
î_Ě༐  Z༐ ༐a_ºX༐༐‡77ºXĈ-༐
ë
¿
 ༐
-
. ༐ ༐ 
༐  Z ༐ 
༐  ¿l༐ ༐ ²
RZ
༐ xS༐ 8  ༐


RZ

༐ B
༐ 8 R༐ y7༐ -R

<༐ z༐ ¿
S༐  -¿ -
l
༐ 
༐ 


S༐

¿
 ༐ ༐ QZ8Zl༐ Z

༐ ༐
-¿
Rl༐  ೴༐
<༐ x༐ R -༐

༐ ༐  Z ༐  -༐ 
༐ Z ༐ 
-¿
Rl༐ ZZ༐ 
༐ .༐  ༐ 
༐  -༐

-¿
R
༐
༐

ZZ༐Q
R
<༐ Û༐ l༐ 
༐ S༐  ༐
༐ ZǞ
 ༐ -

༐ -,༐ 8l༐ ǃRZ༐  ༐   Z S༐ ༐    ༐ 
༐ 
༐ 8༐ - ༐ ༐
¿-R-༐ Z“ ༐ -
༐ .
༐ ༐ R
༐ R༐  ༐ 8
༐ Z

༐  ༐ ,༐ -R
<༐ ]
༐
3Z

༐ ༐ l༐
-, ༐
£,༐ .ZZ༐ ༐ £
༐
༐l
-l༐ l༐
3
ŦZ༐ ༐
,༐ 

༐
È lRZ
<༐

ü #×ü
z༐ .
༐ Ý
༐ 8
lQ
༐ 8༐ ë<༐ ˖
.

༐ ༐ ũ ‹b™ȩω༐  ༐
3“
༐ Z,

༐ 
r
¿Z
༐  l༐¿€¿


༐ ༐ ༐ 8

༐ 
£Z༐༐  ༐  l༐Ŧ“l
Đܶl༐ 
༐
¿ ¿  Z༐  ༐
༐ 

༐ ༐ l༐
l

f༐Ϫ
༐Q
.
༐ ༐ ¿ Z¥
༐ Z ༐

8
.ll༐ 

༐  ‫لۿ‬༐ ༐
ll
༐ ༐ ƞ Ԣ Z S ĉ ༐¿

༐ ༐ Â
l༐ Z ༐


༐ ༐ Ç͏l
༐  Z 
༐ 8l -
༐ QĂ8Z
༐ €-༐ .  ༐ £
༐ 
 
༐ ༐
,- R
༐ ༐
༐-,༐8
༐
R
<༐
Û༐ Z S༐ .
༐ ༐


༐ ༐ 8
 l༐ 
ZS༐ ༐
“
༐ Z,

S༐
8l Q

༐ 
༐ ༐ ༐ .

༐
¿Z༐  ༐ . ༐  -¿ 

༐ .  ༐ ,
༐ lϐ¿lQʻZ༐
ZZ
༐ 
༐ Ö ༐ ߈༐ ༐ 
༐ ƞlt[ ༐ < ĉ ༐]


༐ . ༐ ̹༐
Q
༐  l༐
Z
Z༐ Q8 
༐ ,༐  ༐  Z

༐  ༐
, ༐ £
༐ ༐ ,QlZ༐
  R ༐ 
༐
¿
-
༐ ,༐ Â

༐ Q
Z 

<༐ ]
༐ 
, ༐ ༐
¿,l༐ ༐

༐ ¿Z

༐ ༐ Q8 ༐
¿

༐ƠĐĜ༐
༐ 
£ ͗Ĝ༐ ༐ 
༐ ¿,Z༐ ,t
༐ ༐ 
༐

Z S༐.  ༐
༐
Ý
༐
༐
༐
 ༐༐.  ༐Z ༐
༐
-“
༐. ×
R༐8


༐ Á<
<S༐
༐ ༐༐
t£} ,༐
༐ 
༐Z  ༐

༐ ¿Z༐ t
༐ ,༐ ຂ
Z༐ .զ ༐ 
༐
-
༐ Q
Z Ě<༐ z༐  l༐ 3

༐ ༐ 3Zl༐
îXXĮ Ŷ Ě S ༐ 
༐ ¿Z༐ ,tϐ༐  

༐ . ༐  l༐ t
༐ ༐  Z༐
--
<༐
z༐ 
,༐ 
༐ 
Z
S༐ £
༐ ༐ ༐
t[ ༐ 
༐ ,༐  ŧ

༐ ༐ Q
Z༐ . ༐
dRZ༐ Q
Z 

༐༐ . ༐ -RRZZ༐¿
¿
RZ༐
П
 
༐༐ 
༐ 
Z <༐ X
Z
༐ ༐
༐Ֆ
༐ ¿
S༐ ZB
༐ Z
S༐ 
༐ ZZ
༐ ƞRr
,t,Zdž ƴ ༐њ£,Z
༐ ༐
༐

 ༐£¿
·༐ ༐.  ༐-༐
༐༐
t,-¿Z
S༐l༐ZZl༐
ƞ8}tZ <ĉ༐
 UYП  b༐

Ϫ
༐ Z ༐ 
༐ l༐ ༐ ༐ M
J༐ ༐ ,༐  Z
༐  ༐ 
༐ ¿,Z༐ à
Æ༐ £
༐

 ༐ ༐Q8£ ༐༐M༐¿ ༐ ༐
༐ M[.Q
༐ ༐
༐¿
¿
r
6Z༐  ༐ £
༐ ¿Z
༐  , ༐  l༐
àī M༐ M༐ ༐ 
༐ ¿Z༐ à
<༐
]
༐
Z£Q
༐



༐ ༐
༐. ༐8
-
༐
¿
༐ ༐
༐
 ༐ ༐Q8ŭ
 ༐ ༐ 
༐ 
༐8
-༐. ༐

¿2༐ £ ༐
༐ ¿£,Z༐ à
f༐ ੒༐ 
༐
 ༐
༐ Q8 ༐ ༐ 
༐ 
༐ 8
-༐ 
༐ Z
༐ ')  ༐ 
༐ ¿3J༐ à
>༐ 
༐
2


༐ ༐
༐ M༐༐
àƗ M༐M༐-M༐8
༐ 
l༐ ༐
-JZ
༐
£ ༐ ༐ ༐ £
༐ M༐ M༐ îΟຕɢ༐ >༐ 
2 lÆ༐ 
༐Q
Z M༐Z

༐ ༐- 
༐
P ð ூᛦ ୮ĈΟĚ༐£ ༐ ༐ ༐ 2༐ ŰM༐M༐༐é

༐M
J
<༐ z༐
༐ M༐
M༐ Q
Z
༐ 

༐  ༐ 
༐
à} M༐ MÆ༐ 
༐
¿
B
༐ ༐ ༐ ¿
£Q
༐
M
Z ƴ༐ ༐ 
༐
Q


༐ 
l>༐ ༐
Q
༐M
Z
<༐ Ѱ6‘>༐   ༐ ༐  
༐

Ý Æ༐
༐༐¿
£Q
༐6àZ༐M
Z༐༐Z
༐༐
Ql༐6àJ༐M
×
Z<༐ ƤZ

༐༐

 ༐8
Q

༐
༐༐¿
Ql༐6àZ༐M
JÆ༐༐༐ -¿

r

 ༐,
༐༐
Q
༐6àJ༐ M
Z <༐
Ʀ 6
༐ £
- Z M༐ .
༐

8Z

༐  ༐ ¿ Z‘
༐ J ༐ 8
 
༐ 
༐

 
༐ ༐Q8 ༐ ༐ ༐ ¿ Z‘
༐ 8
-༐.

༐
2-
<༐ ]
༐¿Z
༐ ༐
¿ J‘, ༐ 
༐
༐¿Z
༐ -Z༐ ༐
༐¿Z
༐ ༐ .  ༐ 
༐
J
,Z༐ Q8ŭ
 
༐ ༐ ༐ ¿J
Ɨ¿ Z‘
༐ Z ༐ 8l-༐ B
༐ ¿Z,
<༐ x༐ Z ༐ 8
-༐ 
༐
༐ £ ༐
8
༐ɡ¿ Z‘
dž༐༐
༐¿Z
༐ ༐l
༐.
༐
༐Q8 
༐ ༐
༐
Z
༐
Ql ༐Bl༐ ¿J
༐ ༐  ༐  Z

༐  ༐ 
༐ ¿Zl༐ ༐ 

<༐ x
༐ 
༐
Z
༐
!༐ - 2༐ Ql 
༐ 
༐ -66ZZM༐ ¿
¿
6J༐  ༐
, ༐  2Æ༐ ༐ ŭ
£2
M༐ 
༐

Èl༐. ༐  2༐
Z
༐ Q
£ Æ༐ 2༐ -M༐

M༐  ༐ 
༐¿Z
༐
༐¿ Z‘ Æ༐
༐
Ç౸2༐ 38 QlÆ༐ 
༐ 6

£ ༐  ༐ 82༐ l༐¿Z
༐   ༐
£ l༐ - 
༐ Q
£ <༐ Ϫ

Q2༐ ¿

8Z
>༐ 
༐ 6

༐ ༐ £

l༐ 
-
༐ .ZZ༐ 82༐
Q 2 ij༐86༐
M༐
༐Ý-ZM༐

8J

༐༐ 
༐ Z£
6
<༐
] l༐ ¿౷Z༐¿ ༐é


༐ ࠴ʽᛦ ༐-ખÆ༐༐8l༐-

6
༐8M༐ l༐ 8
l֪ ༐
༐
༐
J,Q
༐¿ 

༐
 ༐ ༐༐¿ J
 ¿
༐
d6¿¿
༐. ༐ ༐ ৭8
Ұ༐
 -¿

£ Æ༐ ༐༐ d6£‘༐.

>༐ ༐
༐ Q
༐8M༐
༐ ZJ . ༐
Z  ̒༐
Ꮤᛦ
ூᛦ ׄΟ༐ ூᛦ *A 0A (A а bja•Ğ༐Ĭ༐
‫ػ‬੢
.

̒༐
ׄΟ༐ ூᛦ  l༐Ç͏22
༐ ༐
༐
!Q
༐ 2
༐ ༐
༐ M༐ ,༐
à£r
M༐M>༐ Οࡋ༐ ˫ Ο߉ࡌ༐
ð u֦ B

༐ ༐ 
༐ Z

༐
¿
-
༐  6 ༐ .  ༐ 
༐Z ༐ Q
Z
<༐
*A ூᛦ :ðĈ
༐ ð ூᛦ


༐
(A ூᛦ


Ɨ ¿Z༐ 
ǃ
༐

²
༐ 
Q
༐ 

༐ 
༐ ¿6
༐ 6-82
Æ༐ 
༐ -

 ༐ ༐ (A 
༐  ༐ ༐
2¿ !Z༐
£l

ƴ༐ ༐ Ơ͗ܶQ

༐ 6༐  ༐ (A 
༐ l༐ 8l.

Æ༐ 
Ç
༐ 
༐ b7Ӧ‫ ؠ‬Ŷ༐
ˡ-û༐Ĉ༐M» Ɗ༐  Ā༐๝l
ĝ} ¿£!Z༐ 
̞
£
༐ ༐ - ‫ށ‬£༐ Z

l
༐ ʻ༐  2༐
à¿

༐
8M༐ Ҥ6-8»‫ށ‬༐ 8
£.ll༐ ԓ ༐ %M z༐ 


༐ 6£
>༐ µອ l3 ༐ ༐ ɤ༐ x Ɍ
ອ ‫ד‬ေ
Ƴ?B"Nေ؎ g'%ေ"
ေ /[`ေ ᕱᛦ :Bnjေ­/
ေ ę<ေ͕ 7)ေÉ
"ေBေ4"B:" ေ
пэa༐ • °( П άͨLП *(zdП ċ°П ŠП däП

½OW–iæ  hေ ±º­iy¯!  † º\O–æ ]zâoo}\{F Hº²æ §oæ s …O²Nãäy9?æ

’
ĭ
Ʌ

4NN(Bေ  ±Iေ
G'ƥေ 4NN(Bေ g 'Úေ
4BʨN
ေ   NjÚb  'Úေ
Ǭ
BေBQBäNေNN(B-
ေ
C
N­ေ
Bေ(k
k-ေ  g ÚÚb g IÚေ
'ƥေƼµÞေ  z Úေ
I ƥေƼ෵Þေ  g Ѧေ
hࣹေ ޵Þေ Ĕ ѥ z Úေ
V</NäN(B"
rေä:ေ
uEB/Nေ Gေ
V</NN(B
èေ :ေ

("<N(Bေေ Ѧhေ
t­N/NN(B-
ေ   'Úေ

-༐ ¿
༐ -ZZ-
?
༐ ༐ Z ༐ ¿ ༐ 

༐ ¿6
༐ 8༐ ༐
l

༐ ༐ ඾ᛦ 8༐ îy7ȩ༐



Ĉ༐-û Ě༐ ༐༐ -?
Z༐ ¿

 ༐༐


}¿?Z༐ 
Ȳ
༐༐ y༐ 8
.ŭ


<༐ x༐ :

༐-༐ ༐ üd<༐ ɤ <aø༐

༐
༐ZZ. і༐

?༐ î -ಗ Ě ༐
²v༐

༐ î༐B༐ ༐Ĉ༐-༐û Ě༐ С֦

7<ø‹b༐ ^ð (A î 8
.


࠰༐ ^ð ð î - © ༐
]D8Z
༐ b <‹༐ :

༐ :Z6

༐  ༐(A   ༐8
.


༐ ༐༐ :l༐  ༐ ೵

༐ ?¿

༐ D
༐
² <༐
²
:
Z༐
ʫ¿
-2Z༐ ¿262
༐ Dă༐ :Z8Z
༐ ༐ 
༐ 8

:༐ ༐
-2
6l-
?༐ ༐ 8

l༐ ༐ Z

ƿɍǕƿ ȩ ij ༐
༐  
༐ ༐ ༐ ༐  2

l༐ .ǧZ༐

¿2༐༐
༐6༐
d6
<༐X--༐-
? 
༐6

༐༐
༐
Z

༐
6
༐ ¿
-༐ 
༐ -

6
-
༐ ༐ ࠴ᛦ ?༐ ˌ _ ༐-౺ܷ<༐ Ž
ů
༐ -
 
༐ .ZZ༐
ZZ.༐ 
༐ -
D
6
-
༐ ༐ Ɉ7<7_༐ -ങ<ƿȩ༐Ë༐ ¿Z༐ Z

·༐ ༐ ¿Z༐ ¿ ༐

é


༐ ¿
༐ -༐
d6Z༐ ༐ _༐ -ੴ༐ 
༐ 


༐

<༐ Ë༐ lZ
༐
ZD

.
Æ༐ ༐
༐
d6Z༐ ༐ ૺࣞ༐ ༐ 
༐ 8
B ༐

 ? ༐ 
༐ 
¿8Z2༐ ༐

Z <༐]B ༐ 2༐

༐Z ༐༐ ԓ B Ĉ--û ·༐ ༐ ũ 77༐B Ĉ-û ·༐ :

༐ ࠴ᛦ ூᛦ
ʵͮ77༐-౹ܷ༐
༐༐ 6¿¿
༐ Z-༐ î b ༐B Ĉ-û༐ ອ ™༐ ^ð y7ͥƿ༐ -Ě <༐ ]


·༐ 
༐ l༐
༐È6༐
d6
༐༐6
༐
༐

-D
༐
༐༐
ʫ

:
<༐
˝ j ༐ X<༐ ë
༐ ‫ޥ‬Ѩƿƿ༐ 
༐ 

8
༐༐ Z.[
·༐ ¿8Z
༐¿Z
¿
༐  ༐ Ù
༐
86Z?༐ -༐ ༐
༐ .[
ZZ༐ ƶ
Z ༐ D༐ 2-B
і༐ Ö²
༐ ¿?D8Z
༐
¿ZD
¿

༐ :
༐8

༐

 
༐༐6༐ Z

༐
¿༐2-¿Z


·༐
—

༐6
༐༐

-
Z༐


༐ ༐  2༐ Z

༐ :2༐ 8

༐ Z-

༐ 2Z-?
<༐ ËZD.
༐ 
༐

:
l༐ 2DZ༐ ҏ༐ ¿

 ·༐ ༐ 
:
༐ 
DZ ༐ -
D

6÷

༐ ༐ 8
༐
B
<༐ ]
༐¿Z
¿
༐ 
༐.༐?B
༐
༐¿Z
༐. ༐--
2
༐༐ Z×
¿

༐
༐༐-6 ༐6

༐-

6 ༐?Z༐༐
༐ Z

༐6
< dž ༐] 
༐
༐
༐༐
82༐
6
·༐ ༐8
༐
6
·༐86༐ 2༐ ƨǚl༐༐?
༐ --༐ ¿ZDÙ
¿
༐
6Z༐
Z
༐ 8
༐ .
༐ ༐ 
༐ Z-
༐ ༐  
༐ ?Z·༐

¿
DZŅ༐ .
༐ [
¿
༐
eZ

·༐༐ Z

}‫׸‬-
Z༐

Z
·༐ 
༐?༐8
༐—6 2༐8༐<ƿ‫ ݰ‬Ǖ‫ ݰ‬Г༐
ß ༐
Z
༐ 
༐ ༐ 
༐ :Z:
༐
68—
༐ ༐ 2d6

༐ 


༐ ༐ 
༐
༐ ળ ǕǬź‫ڵ‬༐ ] (༐H:༐༐ %
:༐ ༐ ༐ H
༐

༐ H༐ ༐  ༐ 5
/H༐


S༐ %/༐ ༐ :
༐ ༐ DF༐ ༐  ༐
D /(
༐༐
 
༐༐ ༐H༐


(
<༐x༐
HH༐
/H༐

:༐



༐ /H༐ 
༐ 5d/HH༐ ༐.༐H ༐(
༐



S༐ .༐ 

:༐



(
S༐༐5༐

:༐༐༐
༐
5

<༐ ]༐༐F (༐%

Ư
H/༐ :H/
༐ ༐  ༐ H༐



S༐ H༐ 5D
/āF
༐ ༐ d/༐
.  ༐ ༐ :༐ ༐
/༐༐ ༐H༐



S༐ S༐ %༐ 
༐༐ ‚
༐༐
༐H ༐༐


}༐%/S༐༐F ༐H/H༐༐Ϊ

H༐



༐༐ ༐
༐༐H
༐d/H%/<༐] 
5༐( 
༐ :༐
%5༐
%༐༐ ༐!H༐%༐ß i༐ȩƙ༐
x༐ H༐ LJʬ ǘ ༐F
༐ %
:
༐ D༐ /5
༐ /༐ ༐ ‹77ÍX༐

༐ :Hě%H༐ ༐ ೐ᛦ Ŋ‘S༐ ’¥H!S༐ Ƥ{!<༐ ] 
༐
/༐ “
༐
Hտ༐
(
 ༐ F༐ ;:༐  ༐ 
༐ ༐ Ĕ ༐ /%༐ /
<༐ Ϙ HǬźϭ༐ 5
༐ ༐ 
༐
/
5༐༐F ༐ 


༐ÇēF
༐%
:༐༐B ༐
F

[D/5༐
;/:
༐༐ H

μ}H༐
5H
<༐

º@ ü ü ü

²


༐ ༐ H

༐ ༐ 
༐ ༐ H༐ %༐ :/༐ ༐ ༐ DH ༐
%/༐఼ցƠ༐༐/༐༐༐HH༐  ༐༐ H༐
༐%/ ༐%/༐
%༐ HH

<༐ ] 
༐ ༐ ༐ :༐
/
༐ H༐ 
d/5
༐ ༐

:
༐  ༐ F/H{༐ ༐ ༐  ༐ /%༐ 
 5f༐ üHH

༐ .HH༐
/༐༐H(:༐/༐ . ༐༐:HFD ༐ ༐
༐

༐༐ ༐
H

༐.  ༐ F
༐  ༐ ༐༐ HBH༐ 
S༐ F/HH༐
/༐ 
<༐
] ༐ 55H༐  
༐ ༐ H

༐ 5HH

༐ ༐ F ༐ 
༐ ༐ HD

Ȝ‚
%(༐/༐H
༐ D
༐%(༐B.༐༐D༐
!%H༐༐༐.
༐
/‚
༐ %༐ ˖d/HǬȩ༐ ༐ ] 
Ǭƿ༐ 
༐ H༐ 
༐ y‹_ͯ༐ D༐ y‹Ē_S༐ S༐ ༐
FHS༐%༐ƤHH/ǬǬ༐༐ ˜

 Sިƙ༐. ༐5༐F༐ ༐HH. ༐H/


ʶ༐
Á y © ༐ ß
:༐
/༐ 
༐  ༐ .!༐  ༐ H༐ Ǻӛ‫  ת‬༐ 
༐ !༐  ༐

༐  :༐FDH༐ Ái5 i S ༐F ༐ ÖÂF:༐D ʐ © i ༐
Ӑ˜©༐ x༐  ༐ Ö5Çē5:༐  ʐ༐ FHH༐
/༐
༐H5
༐ ༐ //‚
HS༐ . 
༐  (༐ H

༐HD
༐D༐F ༐ ÖÂ:༐ Dʐ༐ .HH༐%༐H༐
༐
/S༐(
/H ༐༐༐H༐/ S༐
HDƗ ༐ H

<༐
Áώ©༐ ÛT ༐ 
༐ H
༐ ༐  ༐ Ö(‫۝‬ˢ:༐ ʐ༐ D༐  ༐ ༐  ༐
ÖÂ:༐  5f ʐ ༐z༐
༐B.༐ ༐ HH༐ H༐ H


༐ ༐༐
/
༐༐XÛň S༐ ²৹఻༐ S༐ D༐¯ňؑS༐.  S༐HB༐ ༐
༐XÛ৔ ƈ S༐ ²Ûত ƈ S༐ ༐
Û¯ƈS༐༐༐ ༐ HD

༐
//<༐] 
༐ 

༐༐%༐/༐H‚
HHH༐ ༐ XÛS༐ ²Į û S༐ D༐ ¯ຯͦF
༐ .  ༐ ༐ ༐ HBH༐ ༐ /༐
 ༐ 5༐5/༐༐έ‫ מ‬S༐ ༐X‫ ࠶מ‬i༐ ǴD

 ༐༐/ā༐ / ༐༐

[H༐
HD

S༐  5S༐ /
༐ 
༐ 
༐ < ƙ ̎༐ ऎѩH%5/༐ Åອ F/
༐
H
༐
༐%5༐:T‘༐ %༐:/༐༐  ༐.༐༐༐ ༐ HD

S༐ D༐
/
༐༐HDB༐
H
༐ /
༐
/H༐/༐ ༐Ć/༐༐HH

<༐
ęN੢ •Ë (€ æП ͂ēLП *(zdП /П æП  däП

] 
༐ ༐ Â
༐ ༐ H༐ ༐  ༐ HH. ༐ 
%H༐   
༐ ༐
 ѓ༐
Á b Ě ༐] ༐ :༐ ༐ ༐ ÖÇē:༐ ʐ༐.HH༐:H༐༐DB༐
H‚
༐%/
༐༐ ༐/༐༐H༐T༐༐H5D༐%༐ ༐5H
༐ক/<༐
z༐  ༐ %ԇ5༐ ༐ H5༐ T®༐  ༐ HHH༐ ‫
ف‬༐ ༐
/༐ ༐ 
H༐
.HH༐ :༐༐ %.༐༐%H/5༐ 5<༐ Ö˵/ʐ༐ Á
༐ X ༐ y৑Ȧ©༐ Hԇ༐༐
HBH༐༐%༐H [H༐%/
༐༐ ༐/༐༐//
༐Tf༐
ÁaĚ༐ ] ༐˵/༐H༐ԇ ༐ ༐ɡÂ:༐ʐ༐ .HHS༐%༐ ༐
༐
BS༐ B༐ ༐ ༐ B༐ H༐ %/
༐ ༐  ༐ ༐ ༐  ༐ ༐
T<༐
îȫIJĞ༐ Ƥ
༐ %/%%H
༐ .HH༐ ༐ ༐  ༐ H
༐ D༐ %༐H
༐ ༐  ༐
/%S༐
H ༐ ༐://<༐ ²༐H

༐༐ ༐ 
༐S༐/
 ༐
‹bš7༐ H

༐

S༐ 
H
༐  ༐ 
༐ ༐  5༐ HH. ༐ 

༐ ༐ ༐ ༐

 ༐/
ʶ༐ ¯ ň༐ S༐ ¯ ňĮS༐Ûň S༐ XÛS༐XÛň༐ ༐Ąň༐ f༐
̬‡Ě༐ ] ༐   ༐ ༐ 
༐ ༐  ༐ H

༐ ă༐  ༐ H
༐

༐ /༐

%H༐

༐H ༐༐/<༐
z༐
༐༐༐ H‘༐ ༐ ༐ ༐H༐ ༐ ༐ 5«༐/%༐.HH༐༐

༐ ༐Â:༐ ༐/ ༐ ༐:
༐:H ༐ H[HS༐ ༐ ༐ ༐
 /༐༐ ༐:
༐:H ༐d/H༐T
༐ ༐ ༐.༐
༐ :H ༐ %༐ ༐ H ༐ <༐ üHH

༐ Â
༐ ༐ H
༐ %༐ 
༐
. ༐
H༐}༐:H 
༐༐DHS༐
༐
/༐Â
༐༐ ༐H
༐
༐/
༐༐༐/༐༐‫ݔ‬༐<༐ ƤHH/ƳƳ༐H
༐
%
༐
༐Â༐.  ༐
.
༐ D༐HH

༐༐

༐༐/༐%.༐༐/༐
H༐ ༐ ༐
/༐ ༐  5༐ H

༐
༐ . ༐  ༐ H༐ 
༐T
༐ ༐ %Ď
%༐%༐
༐H
S༐  /
༐H ༐༐ ༐༐༐༐:/H༐ ‚
<༐] 
༐༐.
༐/༐

%H༐༐ ༐D༐༐H /H༐
B
༐ H ༐  ༐ ƍH༐ H
༐ ༐ ༐ «༐ /%<༐ üHH

༐ ༐ DH
༐
/༐ %.༐  ༐ /
༐ ༐ ༐ HT༐ ༐ ༐ :H }.:༐ /%༐ ༐  ༐ H

༐
%H༐
༐༐༐. ༐ ༐ HTS༐༐ ༐Ɇ
༐༐B
༐.HH༐
/Hi༐Ĭ༐

ÍË>t9
ü  5ü

ƤH


༐༐HෳH༐/
༐༐HH༐/
S༐༐ ༐

:༐
༐ ༐༐ y7ŕƙ༐  Ӂ ༐༐༐/༐༐y༐   Ӏ ༐༐b a77ºXÆ༐
 ༐༐ ༐ HD

༐
<༐] 
༐༐.D
༐
%H
༐%༐ ༐H

༐
.B༐༐ü<༐ ’%/ ༐ Áb‹‹‡Ě S༐ . ༐
.༐ ༐ŒcDЮ
༐H.༐༐H5H༐
/༐ 
༐

ƍf༐ x༐ 
༐ /
༐ %༐ ༐ %.༐
/D༐ ༐
:H/༐/:S༐༐/
༐
/H༐%༐B༐/ ༐
/
༐
/H

༐ ༐
༐༐༐:
.
༐ ༐ <༐Œ /༐ y < b•༐
.
༐/:
༐
༐ :H/༐ 

:༐ 
༐ ༐ /༐ ༐ F/༐ ༐ X ༐ ƤH


༐
ĒĒ‡7S 77ƽ7·༐D༐77‹7<༐’ ༐

:༐
༐H༐༐ ༐ H  ༐
DH༐:
༐
 ༐ H༐ ༐ /S༐
 ༐ H
༐ 
/HS༐ D
༐
.༐ ༐ Œ /༐
Ǻᛦ ª3N;$"
ေ (,($"Nေ3ေ 4E
33[ÀÉ$ေ+3E3$"Nေ
 UYП ŰŰП

ĆĒ੢

Ã੢
୫ອ
*

Ã੢
ୀᛦ bᛦ
ࡁᛦ
௚
ᙹᛦ
՚੢
‫ם‬੢
ƨᛦ
‫ލ‬੢
म੢
҆҇੢
‫܌‬੢
Ã੢
ତଥᛦ
‫ݒݑ‬ᛦ
)  )
ࡁᛦ ) ઑ࠳ઑ࠳ᛦ
௑ᛦ
) ઑેࡤᛦ
ལᛦ
ƒᛦ
ࡁᛦ
‫ݔݓ‬ᛦ
) ࠳࠳ई࠳ᛦ
) ઑઑਗ࠳ᛦ

Ē੢

ǁNK­<
ေ  gGgေ ¯ʊ­Lေ <;N"[ေBေ Bေ Ń­(NေŃေ
E
<$"­<
ေ Ń<ေ +<KေKB
ေ

&&&q &&h&Y I&I&% IG&&ေBေŃ­
ေ(Bgေ Ô+iП&
—úƸ››@ £—@fП ǟUέ}ПX›

—›XП*}3Ѝ
„@@ˆǠП „iП ^П +П Þ}3ʫП ĭ@&i—›X?)П ‹ ':੢ Ôŭ3ПÐ3;ÕП •PU—}“П*›––П+@@ˆП ĭ@ω&}iП (P ^ ¿ П

b <a7༐ .

༐ !!༐ ༐


༐ !!H༐ H

༐ ă༐ Ͻ:
<༐ z  ༐
༐w!
༐  w༐
 ༐ ༐


!
༐༐}༐ů2H
༐
༐!H༐. ༐2
 ༐

!9
g༐
Ϳ
!༐ H!

༐7 ɵ a7༐ 
༐༐ 
༐
H2H༐2

:S༐%༐༐

%H
༐w ®༐
 !༐

༐ĒĒͯ7༐ ċϧ
TÕ༐༐¾[H
༐ H¾Ǽ༐77‹7<༐ z༐
®༐

2·༐༐
!
༐༐B2༐:
!H༐¾
¾༐༐

: ༐9%

༐༐7؉ ‫ؤ‬7༐༐7ɵ मӫ༐
H¾ʼn®༐ . H2༐ %/H%¾༐!
༐ !
༐ ༐
![H
༐ H

༐ 77‹7>༐ ༐ 
༐
d/:H
g༐ˆǯ
üH
H
Ͻ
༐ %
0

༐ H
༐ 0

༐ !༐
H
:!
༐ 

w9

༐ Ͻ
༐  /
༐ ‚
‘
g༐ Û༐ 
༐ 
༐ !®༐ 
༐ /

༐ ༐ !༐ 
༐/:
༐ H!

༐༐  2༐
‚
:
ೲ
༐
A

༐ 
༐ /9H༐ ༐ w

༐ 


༐ %༐
༐
H

®༐
Bေ 4B:c Ėေ EN"ေ ­ %ေ  B"ေ 
ေ
æ Í ေ
B`
ေ EB((
ေ 
ေ ေ E[%ေ
KB
ေေ
;
ေ K
<ေ<
N;[ေ" Bေ
BေKB
ေ(Bေ7
ေB
ੜေÞေBေŃBħ<[ေ(Bƾ
%ေ
:
;
<rေ Κ
Bေ KB
ေ B<
ေ E<­(
ေ <
ေ
B‫[ؖ‬qေ :" ­ေ Ǟ(­N
ေ B<Kေ Ń<ေ

;"<Ă(BေBေ­B;cB7
ေ;([O
E
cB­c
ေc
B E gေ
ę‫ק‬੢
N5 !6 "O
 & $ '"O /-П (O $'"O

%,: >,?) DH?,O U 0\

52Lw

ಇᛦ 627Gw
ᎊᛦ
528Cw
˃ԓ
ۜອ
Ꮁᛦ d `
w

52Xw

52Vw
཭ᛦ
ಇᛦ 52Tw
ᖺᛦ
ᎋᛦ 52Rw
௉ᛦ

F(2w A%Ww A*Sw A)Kw A&Aw A'2w 9%Ww 5%Sw 5(Kw

5 22 2 e ^ w
ȭK­<
ေ g&ေ ¯?Lေ<
ȿȿ";[ေ;
<ȿ­ȿေ
S
B­<
ေေħL<ħBေKBȿȿ5ȿ ေ<5ħĎÈ
ħ?<<
ေ;B­
ȿgေȏę"
<ေ4 Bø g΢ ࿹ ٢ေ ߛ‫ࢉݖ‬ອ b@xÆ
iП@Æ›XПU¶П ̕@—ˆ  ¿ П

༐
é
༐.  ༐ ༐
༐༐99
༐༐
༐
Q

༐༐ƒÓ
༐  
༐†༐ ༐


B
༐ég༐
Ø
༐ ༐ ༐༐
ˤ
̟༐ 
༐ Q9†༐ @9Q༐`ā
༐  ༐  ༐
༐ ༐ 
༐ T†༐ . 
༐ 
༐ ༐ ༐ 
༐ 

†༐ <༐ Ø
༐ ༐  ༐
̟9༐ ༐ !†༐  †༐ !
༐ ༐ 9༐ ༐ ༐ ֐ņ༐༐ †!༐ 
×
̟!†༐ %༐ 
༐ ༐ ༐ 9༐ ˾±༐ ༐ 
༐ ɬĹ
༐ ̟༐ .ȇ༐ ༐ !ƓЉ༐

Q!9†༐%9%>༐.
༐ ༐ ††@༐†e
༐
Q
༐ ༐ ƒÓ
༐ ༐  †༐ ҢĹ
༐
!༐༐̓9༐@D
༐% ༐!̟༐!
༐༐ †༐9v
༐༐
༐%9%<༐Ø
༐D9@༐
༐ 9༐ 

༐ ! 

༐ . ༐ ŒŨ༐ !.<༐ Ø
༐
ÇȔ†༐ ༐ 
D̟ ༐
༐ 
༐ ༐ 
༐ Q9
༐ 
̟Qƒ༐ ༐ !༐ ఄ༐ ƨņ༐ ༐ Œ 9
༐
ƽ <˜ş >༐ ! ༐༐ 9
༐%༐²

>༐ ]
>༐ !@༐Ø9@
<ƣ‫أ‬༐Ø
༐9Ư
@༐༐
༐
༐ ༐༐ƒ༐
༐ ༐99ā
༐v
@༐༐†9
༐
v
༐ %༐ ༐ 
༐ 9༐ <༐ ƛ
ȇ é༐ !༐ ]  ǻû༐ ༐ ²
‫ੵݑ‬ĹǻŠ༐ ‚
Q
 
༐  ††༐
ÇȔ†༐@༐ 
!<༐ Œ 9†༐ ɤ <˜˜༐  .༐ ƒ༐ ༐ 
༐ 
†!‚
9
༐ D༐.  ༐ !༐ 


@
༐@9Q༐ ༐ ş7Ʋω༐ Ʋ͂ g༐Ʋ‫ٱ‬༐ ༐

༐ %༐!༐
D༐ ༐  @!༐ @༐ ͭ Ū ༐ɒ!ûĮƲŪ˜༐ ²Į û༐ .
༐!༐@
! ༐ .
 ༐

†e଱ †༐ ༐ 
༐ ²Įû༐ ɞ
༐༐ 

༐ %༐ 
༐ T†༐ !B†༐ ༐ 
༐
9Q†̓<༐ ²>༐ ! ༐ ༐ 9!༐  9༐ 

༐ ƒÓ
༐ 9Q༐  
Ư
Q†Ƅ༐ . †༐ 
ď༐  
༐ !༐ %`༐ T
༐ 

!
༐ ༐ %༐ ȑߺ! ༐
!9@ͧ༐ ѢѮ>༐ Ȟӌ%΁>༐ ǽѮɋ༐ Œ
û࠘‫ۼ‬༐ !ɖʭ
༐ 
༐
ÇȔ
 <༐ QǦ %“@!@༐ ༐ Ȥ༐
!Ɠո༐ ֐Ĺ ༐༐ 
༐`ā༐ !༐
9
༐
༐ DŅBȇ༐ %༐%†.༐e D༐
҂T!@༐༐

༐9†༐%༐
<༐
 UåYП )

ŗ ŗ ŗ ŗ ŗ ŗ  r ŗŗ

आ࠲ᛦ
ŗ

Ûŗ


ಾᛦ
*
*
ಾᛦ
ൄອ
࠱ᛦ Öŗ

$ᙨᛦ
୼ᛦ
ᙩᛦ
˂ಸಹອ
ౙᛦ Òŗ
ࡹࡺᛦ 
ౚᛦ
ಾᛦ
ЊЋ֦

ჩᛦ
ௐᛦ
Îŗ


$ ŗ $ŗ $ŗ $×ŗ ¶ qŗ µ ŗ $ŗ $ŗ · ŗ $qŗ

G º ĩ w ŗ

ȭK?ĻLေ  zñ ေ ‫ܛ‬ǔ೦
(Ĝေ  k ေ;ä‫ݥ‬ԣ
ေ <
//x;-)ေ ęေҕk(c
B/Kေ ƪ
ေB˶ҟkေ Ń ေX] ťक ը ေkေ ߷ອ

ʠBÀ/(BေKB//ZေtŃÓေ4Óྗ
k%ေÍSS
ေkေ Í‫ݣ‬ԣϛ
<  ֲ ‫ۇ‬ေ ¨u@ŃÆ¢iП Š?¢П @M¢϶iП@͑ПU}П
¢ϩ?@±@Xi ^ ¿ П

^
༐ 
$


༐=༐=9"Q"༐ =༐"
$
"9
༐ !༐ %

༐
T$L
༐ %༐

".=༐
T$

"༐ Ld9!£= Ƚ༐
Ȟ༐ Ëອ ð ੢ $5
%)&.
Ȟ༐ ฀ອ (ð ˛ !$'.     .
.

༐ ࣇߺອ ༐ "
༐ $
ůғ༐ =9"Q">༐ '. ༐ " L༐ !%=9£
༐ "
$
!£9
W༐ !༐
द W ༐͹W༐ ৶W༐༐ !. Ú
༐ =£!"<༐ ˚d9"=༐  &:A ༐ B=.༐ !༐Ž! ༐ ༐ ¯×
 
Ũņ༐ ˶. ijǻĨ༐ E"༐ !༐ =༐%

༐
Q
༐=༐"
=
"༐=ņ
"=༐=༐
"
༐% ༐=L༐!" ༐=༐"
༐࠳༐=༐༐"
༐ ̤༐
".=B<༐ ƛ
༐

༐
"=༐ !༐ "=༐ ".=༐ =9$༐ .  ༐ "༐
"
༐ =
༐ =༐ "
༐ ="
༐ =༐ "
༐ ".=༐

d9"=༐ Q
༐!%=Q
<༐
Ż%ǻź༐ Q
༐༐ 
Q
.༐=༐ "
༐$
=
༐EQ=Q
༐ ༐ "
༐

"༐ =r
9"=༐༐ <༐
^
༐ 
!9

"༐ =༐ " L༐ Q=9
༐ 
"Q£༐ =༐ !༐ ༐ =%ņ9
༐ %༐ £
༐
$LL
༐ =༐ Ч=!=9༐    ༐9
"Õ༐ =༐ Ö

"༐ %=$"=<༐Õ ǻŖࢌ ƣϊ༐
’
༐༐Q="
༐༐«"༐$$
 >༐ "
༐    ༐9
"༐E༐ ̤༐༐9 ༐ L༐
) •Y Š(Y П YLП *(z2П /П ¼ķП 2+П

ᖄ࠭ᛦ
GI
I
৵৶ᛦ
৵৶ᛦ

oABGp
ঽাᛦ ໕ฟ༢ᛦ
‫ۂ‬ᛦ
̉ອ I
I ̾χ̟ԓ
ෟᛦ
‫ۃ‬ᛦ
˥ອ
Ԉ֦
ɓອ
रອ 5੢ 5 Ӻ ੢ ়ᛦ ĊI 
I I Ą੢ ়ᛦ I I ੑᛦ
‫ۄ‬ᛦ
ɔອ
ᛦ ᖆᅑᛦ ᖅᅐᛦ

F[Gp
‫ۅ‬ᛦ I
ᙥᛦ I
‫؀‬ଽອ I
ɕອ
੦
ঽাᛦ
5 I ཇᄟ༡ᛦ టফ༢‫ٱ‬ᛦ
/-   I
I
I ąѾ੢ ়ᛦ I ੢ 
I

੢ 5੢ Ċ‰੢  ੢  ੢
',402CI >-@)-;DI !F4+-I "-?8(*4<1I #5I
 I4=I E3.I
9(ABI ऄ઱žᛦ  (  !  I ઱পŸᛦ #6!I
ǁxK­¦
ေ  gေ ë
ေ "
E
ϚB-ìϚ
ေ B¿ေ : ( ေ ¿
ေ njE
(Ăĉ(ေ (̫қŻ"ĉ;N"[ေ `L(L/ေ अ ࠰ Йળᛦ
 ੭Żေ Eä-"
ေ BpBȿ"ေ " Lေ E
ţ(
ËBKLေ `ɳΤေ :
K "ေ Ńေ Ë
ေ (‫
¿ݮ‬ေ Ǐx
ေ ţ
༦‫ݮ‬ħpေ
4Þ߷ေေ-
ေp]njnjေ <bǶ XBฤؓYေ૝ေ4x‫ץ‬ေtŃ"
Ϛေl
 ȓǍေBေë ୴ ѰĽေ

˝Ԛ੢
।ᛦ(WȜ-ԓ
Ӷԓ(W--ԓ
ӷ(˹Ȭԓ
ԉ֦ "I %"&I
Ƞओ˞ą੢
‫ۆ‬ᛦ
ᓼᛦ ˟ġ੢
 I

 I $7:-I &) #'&$()

ᙤᛦ ͝ ġ੢
ӵԓ͞5‰੢
ऱອ
ਰ ͟Ąӽ‰੢ #"I %"&I

ᙦᛦЖ঩ࡣᛦ
଍ᛦ
ኺᛦИবਖ࠯ᛦ
Зবલ࠮ᛦ
˸ԓ
²Kì¦
ေ   gေ ë[Ex(Bäေ ( BţK
ေ Bေ /( BϚK
ေ (­Ϛ;
/ေ Ńေ ௡ອ Kƾ B// gေ tŃ"
Ϛေ l˶‫ז‬ဤ࿺
Ϛ z ɿɿ ੝ေ Ŏ ‫࡬ ݕ‬ອ
¨x—'iП Wǯϊ'ƛϋП 'Æ…›ƣП Ƣ›'i ^ E П

 ༐ 
༐ ༐ 9༐ 9
"༐ ༐ ༐ 
༐ 
Q
༐ 9༐ 
༐ ༐



༐"༐9
༐༐༐Q9
<༐ ²W༐ e
༐
Q༐༐
༐
༐
Q †W༐༐ 
༐9
"༐༐%
Q
༐.  ༐༐
༐
Q
†༐༐
༐   ༐
୪9
g༐
 )ǯ )

ྐಽ၆ಽྐྸఉ๼ᛦ   , ྐಽྸල౶ာఊ๽ᛦేഒఊྐ೶ጔᛦ

ȕᛦ ধধ੆ᛦ၅Ꮟጓᓃᒉᛦ˘ ৩ࠩᛦ ྷᆻᅎ‫ٮ‬ᛦ

Гंःধ‫੆٭‬ᛦ ψω֦ В ࣿࠩᛦ ྷᆻᅎ‫ٯ‬ᛦ
Ȗ੆ࠪᛦ

࿿ቷፕᆻᛦ ቸ፺ᛦ ྷᆼᅏᏐ፺ᅹᒊᛦ

ਔࠫᛦ ੳࠫᛦ

Еধࠪᛦ

Д ਕࠪᛦ

˚ ੲࠨᛦ

ǁK­<5ေ Í 'ϩ É
;
<ေေ<Lnj­ေ( <K
ေေေ`<ҩ("
ေp/ေ|
ေ( <KKေ()(ʊ
ေ

ေKေ;
ေေ"
ေХK­<
ŭေ <
K/ေ :
<
ေ"Q5ေ"
<ေ(
("NKေĎ 5ေ( B<KKေ`BÛĎ
<)ေ
tĎ
<ေƁ­)
<  Խ Խ ࿻ ေ2 .?æ rM&+8CǯW å&²*ǯ ä”ǯƉ+C2Nǯ

Ľອ
ಇᛦ &\
ഷᛦ
ϟϠ֦
ᅚᛦ

-Á  :'(
ཫᛦ
ϡϢ֦
ϣϤ֦
ईອ
ອ

঱ອ
\
ǣອ
āࢃᛦ
ϥϦ֦
Ľອ
‫ې‬ອ
‫ݑ‬ອ
 II84A87a;B5TSa

 
\ \ \ \

1=H8a =Ia H?JZW8Sa 7YR=I;a \<@6<a

[K AX4;8a = S a 4 M MA=87a
ȭNK‫ݤ‬ҟ˴ࣧǶေ  ?ϩ –ǔǍ
("ေ ေ ( Kေ ေnj­`
ߒ­
Ďေ 
Kေေ Ď
ေ ( <KKေ (­<<
"ေ
ေေEေ`</(ĎLေKေt"
<ေl­)
< Խ Խ ࿼ ေȈ Ñ À͕ †M&+8Cǯ &J ǯ Jǯ+C1Nǯ

Û༐

"Q
༐D༐༐   ༐༐   ༐
W༐ 
༐
༐

༐ 9
W༐ 
༐ %༐ ¯BÙ༐ 
ď༐  ༐ ༐ .  ༐ Ö
༐
Q!༐ ༐    ༐9
༐ L9 ༐ ༐ 
Q
༐ 9LQ
༐ Q༐
༐ 
༐ 
༐ Q
༐ ༐ 
༐ 9༐ ༐ 
༐ Q9༐ Q?༐ ༐
   ༐ 9
༐ .  ༐ .9༐ Q
༐ "B
༐ ś
༐ ༐
 ༐ Q
༐ Q!༐
༐ %

༐ 
"
༐ 

༐ 9༐ 
༐ 9 
༐ 

< Õ ƾ ༐ƛ9
ƣƣ༐
Eǯ ƦǦŜ 
ǯ :ǯ 
/
ǯ . ǯ

Ü ǯ  "
ǯ

༐
%
༐
9

v༐༐

༐
é
༐༐Q9༐ 
 ū༐Œ 9
༐y g˜চͥ

֊ g˜ƪ༐ .༐ ༐L9<༐^
༐é


༐%
.

༐ 
༐༐ 
༐9
༐
.༐%
༐!༐༐é

༐v
༐LQ༐༐

༐ LQ
%
༐  ‚
 ༐ 9
༐ !༐ ༐ 


༐ v L༐ 9
༐ 9v༐ 9
༐  9 ༐ v
༐
g༐ x%༐ 9
༐ ༐ LL༐ ༐!
༐
༐ !༐ 
༐༐ 9ŧ


༐ ༐ ÁŒ 9
༐ ֊ <˜ƪģ<༐ x༐ 

v!༐ ༐ !!9༐ !ā
༐ ༐
 
༐ 9
༐༐ 
༐༐
༐ %!༐༐v
༐99
༐ ༐%
L༐ QL༐
%༐’
ຫgЏϤϤ༐


ü  5ü

^
༐ 9!
༐ 9Q༐ ༐ ༐ ༐ 9
༐ ༐ ༐ ϵ༐ ༐ .
ɂઢ͂༐ ༐

༐ >༐ ȊƸ  ༐ 
༐ 
v
L༐ %༐ 
༐ 99 ༐ !¾ʌ


༐༐ 
༐
Ȋ‫
׆‬༐ 9༐
Q ༐!༐ 
༐ ³L>༐ ༐!༐v༐ 
‚

9
༐ !༐ 
9
<༐ ˘é

༐ ¾L¾༐ ă༐ Q“9ĝ༐ !ՏēLL༐ %༐ v

༐
v<༐Œ 9
༐ш <˜ƚ༐ Q
༐


Q
༐¾༐༐¾9!
༐
¾¾Q༐༐9
༐
!>༐ ċǴÏTÕ༐ gaftǯ !U༐ȧL༐ ۲༐ Aǯ !v༐˾ē

༐
!vQ
༐ 9“×

༐ ༐ ˜7ºX<༐ ^ L༐ QLL༐ %v!%
༐ %༐ 9
༐ 
!
༐ ¾༐ ༐

༐%༐༐
༐
g༐ ²9 ༐¾9!
༐

༐༐QQ
༐9  ༐
༐

Q!
༐ 

9
>༐ ༐ 
༐ !v༐ ༐   ༐ «>༐ ༐
9
༐༐<༐ Ǣy <༐ ^
༐
é
༐༐ ༐༐ ༐ 
༐ L9 ༐ 9
༐

Q༐༐Q
༐9
>༐>༐༐>༐ ༐L¾L༐༐ LLvҽ
༐
Q9!<༐ ÿL  é>༐ ^ ƣÒ༐ !༐ Œ9ŕȝû༐ .

༐  ༐ v
༐ ƍ༐ ༐

TL༐ ༐ՏL༐
!<༐
i੢
©੢
։੢
Ւ੢
ͺΓ
௺ອ
 )     , ,
i੢
‫ۈ‬੢ 
, , !,
˕ ˖ ´
H H ´

ׂ੢
) H˗
H

ᗊᛦ
഼ᛦ
޿ᛦ
) ઍઍਐ࠙ᛦ
௘
ᑈ
ӹ੢
 )
ࠉ੢
ı੢
‫ދ‬੢
")
i੢
‫ݐ‬੢
i੢ )
) ) )  ) ) ) ) ) !) $) )
ӳԓ
टठ੢ ྑೀ ๿ఌခෂ၇ುᛦ ‫ ׳‬۵ ‫ ڛ‬W ‫ ׆‬W Ŗ‫ܘ‬੢–ᛦ
©੢
Փ੢
ı੢
mxKR<
ေ  
ϩ 4R
(
ေc
/x/¿xvxϞ࿸ေ
/ေ
ေԙШࣦ˵‫¿׀‬xwေေc

Ëxv
ေnRÐxŶ‫ط‬ေ ေ ϊ,/
œေ/Ðäx(
ေ
K
//%ေ7c/ÐäǕ(

ေ

(¿x(
ေ Kƾ
//ေ ź+ÃߗÐwKေXေ @@Ĉေ
wေ x5ေKƾ
//ေ ź+LjwKေՁေ
h ေtÛ5ေŠRą5ֲֲ ঢ ေऌ ࠹ ੉ᛦ ‡M+8Cǯ ǯ8U$+M+ǯǯ #ĝǯ
3$8 Qǯ
 )Öǯ E9ǯ

x ༐   ༐Q
ƒ D༐ %  ༐ƛ

༐ D༐ ɑ
W ͂ ƹ˰༐ 9D
༐ 9‚
Q༐ D༐ %
༐ 
༐ ༐ D
༐ D ༐ 9̞
༐  ༐ L༐ ༐

‫ڰ‬W༐ %9 W༐ ༐༐ TLW༐ ༐ %༐༐ ƒ

W༐ %༐
9 ༐


༐T
༐༐ 
༐9 ༐
༐L
༐༐
Q
D༐ 9Ũ༐9D@ <༐
^
༐ B
༐ ༐
༐ 9QL༐ Ý༐  9༐ 9L༐༐ ƒ
༐ D༐ 9
༐ ༐
༐ L༐ L༐ ༐7<77‫ؤ‬༐ <W༐༐ 
༐ ༐ ༐ƒÓ
༐ 
Dƒ
༐ 9
༐ ༐%DB<༐ z༐༐
༐ L
༐  ༐ 
༐ ݅ǚ༐Dƒ ༐.

༐ 
༐ ༐ %Dƒ
༐ ȾȾ༐༐
༐ W༐ .  ༐ .D༐
༐ 
9
༐ ༐ 
 <༐ ^ ༐ 
 ༐.༐
T

༐%༐
ϟ

༐ Ś
Wଏະ੯Ĩ༐ D༐
༐ 

༐ %༐ Œ9BŕȝĨԸW༐ . ༐ 9L༐ %D
༐
DȾ
༐༐༐༐
D༐ 
<༐
²9
༐ 9Q༐ ༐ DǼȾ༐ D༐ %
༐ DB
༐ 
9
W༐ D9D༐ ༐
DȾ

༐ ༐  ༐ 9ƒW༐ %༐  L༐ D༐ ༐ 
D༐ 
༐

༐ ÇW༐ 9 ༐ D༐ L<༐ ë.༐ X@ ༐ ˜77W༐ ༐ ſĘĸÇ fŒ tǯ D
༐9
༐༐
 ༐ 9Ű
<ŵȝĨϭ༐ ëXœ˜77༐ ༐ QDǸD%
༐ ༐ ༐ .
༐ D L༐ ༐ QȾ
 Ɗ༐ D༐ ˜[
LL@༐ 9༐ %༐ .
 ༐ ༐ 
 

༐  
W༐  
 

W༐ ༐

 
 

༐ Ⱦ༐ 
D
༐ D༐ ƒ
༐ D༐ ༐ 
༐ D
༐༐༐
༐ W༐ ༐ 
QDW༐ D༐
༐༐ 
} ༐ 9༐ D༐ ༐ 

D9
<༐ ^
༐
˾༐9༐
༐%
༐9
༐%༐
 ༐༐D༐D༐ ͸ȉèèºX༐༐
[ ༐ 9W༐ ༐D༐
˜ɎƪºX༐༐@L༐ 9<༐ëXœ˜}ु ‫ ‡ ؗ‬ऋ ༐
d9
༐༐ QL Ɗ༐༐༐9
༐Έ ༐.L༐
༐༐%L༐
༐༐ ༐༐༐ D ij༐༐@ W༐Ű༐༐.[

9L༐
%B
W༐ LÇēL༐༐ 9Ű <༐
xƒÓ
༐ 
 ༐ ƒÓ༐ ༐ 

༐ 9Ⱦ
༐ ༐ ƒ
༐ Ⱦ9
༐ 

༐ ༐
W༐ %9༐ .  ༐ ༐ ƒ༐ Ⱦ༐

༐༐ %Ƃ
༐ ༐ 
༐ 
Q9༐ ష֬W༐ ༐ 
༐
D༐ ༐  
༐ 9༐ Á
< <༐ 
  
༐ ÁX¯ଲǫ û༐
²X‫©ڸ‬༐ ༐ D༐ L༐ Q
<༐ ^
༐ D༐ ༐ %L༐ L༐ ༐  ༐  
༐
Ⱦ9W༐ ༐ L༐  L༐D༐%༐Q‘
༐ ༐  L༐ ༐9
<༐ Œ༐
༐ 

༐ 
DW༐ ༐  B
༐ ༐
༐
༐ ༐ y77༐ 
9༐ D
W༐ 
D༐
Q
༐ D༐ . ༐ 
༐ D%%
༐ .D
༐ƍ༐ ༐ 
༐ D<༐ ¯ ༐ D༐
D༐༐
QQ
W༐ ༐ ༐D
༐༐
 T
༐ 9༐༐
༐ %
༐ .  ༐


༐༐
༐༐
 ༐ 9༐ƒ.D༐ 
༐9ƒ
༐

༐9D
< Ò༐ Ĭ༐
^
༐ 9
༐ ༐ D
༐ ༐ %
༐ D
༐

D༐ 9Q
༐ ༐ D@༐ 
×

༐ 
L
༐ %༐  ༐  ༐ ˾༐ ༐ L9ŰW༐ ༐ 
ƒDภW༐ ༐ 
༐
9;
<༐ X9‫ڼ‬Q
༐ D༐Ė༐Ό
;Ǹ༐QD%
༐ ༐ 
༐ Ʀ<༐ ²<༐ x<༐ 9@
༐

༐ D
༐ @
Ⱦ༐ ċü
ðƒD
W Õ ༐ ċ༈
ާңW Õ ༐ D@༐ ċü}X༐ ƛÕ༐ ༐ ˶%%
Ư
Û.
Ⱦœȭ༐ Ϝ༐ X@W༐ 
༐ ß%9 ༐ ß
༐ ƛ༐ XDW༐ D@༐
XɆ ༐ Ϝ༐ ’BW༐ 
LðQ
<༐ ^
L༐ D
༐ Q
༐ ༐  ༐ ƍ༐ ༐ ༐
T
༐ . ༐ ð༐
༐ DQL༐ DL༐ ༐  L༐ 9D
༐ %༐ ༐ ༐
@ð༐ Ǹ༐ ő༐ ༐

QL༐ ƒ

D9
k༐^ L༐ ƍ༐ 
༐QL༐ 
D༐ ༐

æ tေąE8ေေŶ 5ေ‫ݥ‬қ5ေေන༗5ေေ n5ေ K]ေ ƽ8-ąေ:]ေ n
ေ ေtEေGÚhေ

µąေ-
ေ 4(Ó-ąေေ௓఩]ေë
( Kąေ]ေn
ေE]E
ေ:55ေE87n
ေေ¿
ေ”8]ƾေေ

ေ4(
ąေźġࠤ eေ'ϩ ‫ٶ‬t8K8-ေGÚhj j g ေën
ေ]
¿ေေn
ေ
]
ေေ]ေ]:ေ¿ေ]ေ
vƽ5:ေ
E]E
ေ ေ
ေ8](5ေ5]-5¿ေ ေK]e šȟϻ࢏ေ
 ša༐ ŠāƖ 
 ǯ :ǯ 
/“
ǯ . ǯ

 ǯ "
ǯ

.
" † ༐D༐ª༐%
༐
D"
༐༐༐"༐
Q
༐ 9
༐


༐
" †༐

." 9"༐ D
%༐    ༐ "
༐ 
"
g༐

(&&@@ü&&ü

ƛòD༐༐༐

"W༐ .  W༐ òB†༐ "༐༐" †༐ ༐ 9ò,"Ȃ>༐
༐༐

ò†À†"༐ 9
"༐ ༐ D"༐ ༐E"†""
"༐ ༐"†" ༐ †ò†"༐
Çg༐ z༐༐†ò༐

">༐" †༐D

"༐9
"༐*ð ༐
T"༐5ǯ
†"Ҿ
,J༐† ††༐9"༐༐ †༐." ༐" †༐
༐Q"
༐༐" 9༐
༐"༐QQ†༐
༐ 9"༐ ༐
† g༐ ϧ J༐

"༐ 
Q
༐ 9«J༐ " ༐ ">༐
༐"
༐9"༐༐.†>༐  9 ?
༐༐ Q
༐%༐"
༐
T
༐
‫܋‬੢ ອ Ԑ༐ 0A *ð 0A ༐ 5A î y <™aĚ༐
. †
༐ Ԑ༐ ༐ "
༐
é
"Q
༐ Q9
༐ ༐" †༐ ?
" ༐ Q"
༐ ੢ ¹੢ ༐ .”ð
༐ *ð " †༐ 
9" ༐ 9
"༐ .  ༐ 
༐ "
༐ Q"
༐ Q
"༐ %༐ "
༐  
༐
D 
༐ 5 A Ø
༐ 
"༐  
༐ રᛦ ອ ) 5A ༐ 
༐ "
༐J༐ 
W༐ ༐
˓

"༐ રᛦ ༐"
༐ "༐"༐༐"
༐ 

"W༐༐"
༐ 
ò
"༐ "༐

"༐ !. Ø
༐ .ð0# ð
!.
-*A
˩ y <™™༐‫ם‬༐

౵ᛦ Ú֦
˔  -* + A
]
༐ % ð0# ð i¾xŒð  Ҭ ༐ Q†༐ %༐ " †༐ 9"༐  ༐"
༐ ĝĝD"քĜ༐ "ā༐ ༐
"
༐ 

"༐ ""༐ ̓༐ ༐ " "༐ wxð ூᛦ "D༐ રᛦ 0A ̓ ij ༐"༐ ༐ "༐ "༐
"
༐

 ༐༐
༐ð
༐
༐d9
༐"
"D༐ †?༐ *A 
༐9"༐Q9†W༐
༐༐
"༐༐"
༐. ༐
֗9" ʶ༐

а༐ɵ <™‡ ɢ༐

.  ༐ Q
༐"
༐

 ༐"
༐
༐9"༐ Q9
༐ .
༐" ༐«†༐ *A ூᛦ ԐĈ஌༐

༐ Ņ
༐༐ "
༐ 
ò
?༐ ༐ " B
༐ ð ༐ [
?༐ ఈᛦ٬ᛦ
,ð %  ð 
ð ༐༐ 

"༐ ༐ Q
༐ %༐"
༐ Q"
༐ †"W༐ * A
"༐.  ༐ %
B.༐ 9W༐ ,༐ ༐ %
༐
T

༐ %༐"
༐ #.
ð $î
#ð ༐  ༐ D༐ D༐ " †༐
T

"D༐ £༐ D
༐ 
Ç
g༐ Ø
༐ v
"༐

Şƨ"༐%"D
༐%༐ÇȔ

"༐ Q
" ,"༐ v
༐Q༐." ༐.
༐D ༐
༐ 9"༐ ༐ "
༐ ༐ Q%
༐ .  ༐
"
༐ “?༐ "
༐ 
9 ༐ ʌ

9
<༐ Ø
༐ 

"༐ "
 " ༐ ༐ W༐ ༐ "9>༐ “༐ ĝƠ༐ 9 ༐ 
༐
? ༐ " ,"༐ ༐ "
༐ 9%"
>༐ 9 ༐ ༐ ༐ W༐ " ,">༐ ༐ 
9 ༐
%
,B.༐༐ ĝ༐D
W༐ 
༐༐ǹB༐"༐†9
༐"
༐%
B.༐ ༐"
༐
99 ༐ 
9W༐ Ʉ༐ "
༐ ༐ 
ĝĝ",༐ ༐ ༐ ༐ "
༐ 
ĝ
†༐ ༐
%%
"༐ %༐ ༐ ༐ Ņ‘
༐ ĝ D
༐ ༐ "
༐ .
DB
༐ ĝ
9༐ ༐
d9༐
9< ŕȝź༐
 )Iǯ

Б
ЖЗ֦

"#  #
ᔢᛦ
քօ
੢
೪ອ  #
ó੢
ˤອ
তᛦ Ӵ֦ ԸԹԺອ9֦ӸӹӺ֦
‡ᛦ
ௐᛦ
ᅛᛦ
৹ᛦ
࠱ᛦ
ķᛦ
ͫ੢
Œອ
ó੢ !! ! ³ԕ ͜
³!! ӈ!!
҂੢
 #

 #
#  # # ଐ ٧੅ ٧ੱᛦ # # # # ੅ ੱ ᛦ   #  #   #

!#  ! " #$ %%& ]^ƒ

ŝ
U͕ ѥ I ေ Ø@
’ G͕ "7
͕ o
ǯ - ’
''͕ ͕ ''͕ @͕ Ź
@
-͕    '͕ ̮͕
@͕ 
̝
@-

P͕ 'ေ 4 
͕ "|
͕ @'
͕ 7 '4[͕ m/͕ ɩ@' ͕ 

@ ͕
' F @
7͕ ͕ )@' 7U͕ '' Ž͕ m͕ ɪ-@' ͕ U
{-@ ͕ '@
•͕   ͕ ' d
͕ '' Ⱥ͕
ɾŮʬ੢ Á
'-͕ -
'͕ "é
'͕ "*
͕ @͕ )@' -
͕ '' Ž͕ ȉ΃Ēေ Ø@' 
͕ ''͕ FU͕
U@'U ͕   ͕'7͕ Ž͕ m ɏ  ͕ $  ͊ Ĝ U͕ ''͕ 7
͕
@'
͕   ͕'̶ v   | ͕ Ȼ͕ mŝ͕
@' 
͕|''͕
͕
@'
͕  ͕'
x   -͕ Ž͕ ʀѽʮ੢ @' -
͕Ĝų''͕GĦ

@͕ U͕ '7
͕ 
@'
͕   ͕ 'ʤʾ7͕  Ž͕ ȉ{Ʊေ Ø@' 
͕ |''͕ G U@͕ 
͕   ĥ
'@͕ 
@'
͕   ͕ '
   -|͕  n͕ /-
@͕ $ ! Ô À ͌ ͕ ¾ ǻ þ ͕ r"Xǯ %æ$Ǘ
$*ǯ çǯ
ó);
ǘǯ
*3$*$*3ǯ *#_l d ô ǯ_XǯH1ǯ"1ǯ  *#Ņǯ%X$3ƛ¶ ÈÐÌ¥ǯ†£*#ǯH#VNǯŞƜ) Ǣǯ$ǯ
"kǯ% *Xuǯ*Ɗ1 N ǯ

Œ Ô†༐ y <˜Ɏ༐ Q
༐ †༐
B༐%
B.༐Qś
༐༐QÔ༐ 
>༐ ††༐

. ༐ ÔÔ༐ 
༐ ӽ7ө
༐ Q
༐ ༐ ÇȔ

༐ ÔÔ ༐ 
ಜ༐>༐
ś
༐ ༐

༐ B
<ЏƹŖ༐z༐༐
༐༐

༐ÔQ
༐ ༐

Ô

༐ %༐ 
༐ .B
ɫƹǐ༐  ༐ 
༐ [ś
༐ ċ༐ 

༐

  Õ༐ ༐ 
༐ %

༐ ༐ ƍ
ś༐ 
༐ Ô
༐ %༐ 
༐


༐
Á

༐
Â
ģ༐ ༐Ô༐༐
Q
༐ś༐Q༐
༐†

<༐x༐ 
༐


Ô
W༐ 
༐ 

༐ 
  ༐ ༐ Ņ༐ 
༐ >༐ ༐ ༐
 .༐ %༐ ² ȥ༐ 
Ô

༐ ༐ X ༐ 
༐ ༐ Œ Ô†༐ y <˜Ū<༐ ^
༐
ÇȔ

༐
Â
༐༐ī༐༐ࢍ༐Qś
༐༐%
B.༐༐†
Q
༐

×
Ô
༐
༐ Ô
༐ හ༐
༐ ༐༐Œ Ô
༐ b <˜•<ŵƹŷ༐˛༐ ༐ ɒ༐Џƹϊ༐

Ô
༐%
B.༐༐ ݅Ĺ༐ 
༐ 7k77Ū༐<༐  B༐ ༐ ༐ 

Ô
༐
༐ ༐ ƟIJ77ĵX༐ ༐ Ô༐ 
༐ QÔ
༐ %Ô
༐ ༐ ^%
Ѻ <•༐ .

༐ 
༐ %r
ɚ
Q
༐
B༐Qś
༐༐

༐

༐
༐
T
༐༐ŻĽ༐Ĉ ༐Á<  < ࠮ < ༐
^

༐ ༐ 
༐ Q
༐  ༐
༐ .༐ 
༐   ༐%
B.༐  Ô
༐ ༐

༐ 

༐ 

༐ ༐ 
༐ Ô%—
W༐ <
<>༐ ċɚÔQ
Õ༐ ༐ γ

༐
%
B.Õ༐ Q༐ ċ
ś༐ %
B.< Õ ༐ ¶༐    ༐ 
༐ ɚ༐ ༐
ś>༐ 
༐ 
ś
༐ ༐ ༐ 
༐ 

༐ ༐ ༐ ༐ %
༐
 b_ԓ PR 
6Rǯ Rǃü:ǯ H¯
ǯ . ǯ
H ǯ "Hǯ

থ٨ࠦᛦ ࣽ٩યᛦ ࣽ٪ੱᛦ ࣽ٧ਓᛦ ࣽ٧দᛦ ࣼ٨ࠦᛦ

åå߷੢
ł੢
ੱᛦ
Γ
੅ᛦ
Γ

$
১ᛦ
ᐷᛦ
$
$
$

$

$
$

$
ࣻᛦ
࿾Џ
ࠦ٫યᛦ

ࠦ٧ੱᛦ Б
ࣺᛦ
ࠦ٨੅ᛦ
$
ࠦ٨ਓᛦ
ඹᛦ
යᛦ
ࠦ٧১ᛦ л
රᛦ
$
ࠦ٨দᛦ ‫ٺ‬੢

দࠦࠦᛦ দ੅ࠧᛦ ২ࠦࠦᛦ ২੅ࠦᛦ ਓࠦࠦᛦ ੅ࠦࠦᛦ ੱࠦࠦᛦ ઐࠦࠦᛦ

&),'&(%*+(&, &(

:É͕ è ?1ϩ Út 
t ͕ *
’ G͕ ¶͕ 
ê͕ V͕ -͕ ˟ 
͕ 
É[͕
Ť x-͕ *
’ G͕ 
͕ Ê È͕ ͕ ϩ ͕ ͕ 9Ò͕ Ì  ' !͕ /
͕ $ ! į À Dz͕ ¾ ǹ ő ͕ ‡"Xǯ%à4ǩ
å88ǯƗ&‚ǯ õ); 88ǯ
33ǯ  #_Ʃ¾¼ǯ _Xǯ
eǯ "ǯ  *#ņǯ%X4c3+ǯ ‹ 'ėԗ੢ ‡£#ǯ

#VĈǯPళ)ඃ෫ອ $;Gǯ"lǯ¸% X s ǯi Qǯ



༐ £

༐ ÁŒ 9
༐ Œ ,ÑŃŸ Ňǯ 
༐ Qv༐ ༐̓༐ ༐é
ȇ༐ ÇȔââ༐

%༐ 

9
༐ ༐ 
d9
༐ ༐ 
༐ 
༐ Q â༐ ÁŒ 9
༐ y <Ɵঐy © Ɗ ༐༐
£༐ ༐ .
༐ ༐ 9
 ༐ 
 ༐   â༐ . ༐ 

9
༐
༐
d9
༐༐ .༐༐Œ 9
༐y <ƟϏ˜༐£༐y <™‡<༐Ĭ༐
࠵༐ 8͕ -t͕¶͕ ͕
 
͕͕ 
UçÊU ͕͕"
͕
͕ *Ì͕ʛ

˻ ¿¿ dz ¾¿Œ͕ Gt ͕
"U͕ ͕
-͕ *
͕͕ ͕ *˨ ̯͕  U ͕ƃ˼͕͕ 
  ͕

ʮ [͕ ɫ-͕
͕ G͕ -a͕ ͕ G͕ 
-S
3͕ 
͕ 6-
Ì͕ 
-͕
'͕ ¶͕
͕ HûÒǧ À ͕ .͕ ¶͕ ͕
 ͕ "X͕ -͕ *
’ G͕ "Ĝ͕ ¶͕ ĝ͕ ͕   ͕  ̘ U -͕ ¶͕ U ͕
- ’3͕ Rߥອ  
 U͕Ê
3͕ ͕

  *͕G-͕͕͕"͕*A͕

Ì͕
Ë
ź͕
ƋƋ[͕
  ##O )

( ‹7Šfó
%

෿ອ
ਭ੢%
Éອ
Éອ
தອ ু‫׽‬ও‫׾‬ອ
ਪ੢
%
"%
%

͹Γ
%
$%

ഹອ

ਫ੢
ਬ੢
ƟǞ

[Ǟ
% %  %

™)=?(*ˆ!@*(ó ̖ূঔ̻ອ
ĆS
U͕  Šϩ ŗ
_ G#͕ "U͕ ͕ ͕ S ͕ ͕ 
aS
͕ Å
͕ 
çÊ͕ S

͕
 ͕ 
͕ 'ÏĂϩ ˧+ေ Ġ͕ S

͕ /
͕ ɖ4³
! ī ī ǒ ͕ ȉ Ñ ő͕ rƳ·4+8Cǯ Ŷǯ 
++M+ǯ Ƙǯ
> #ǯ
38 ÃQǯ

5*vøĚ  ‚ 1 ϩ k ™ åLaº­}[æ Z­ÎMO‘d Ç)æ   œ æ ʃչҲ˺֫ԕ¤੢ ۚ֬Ň‫ו˹ܺڻ‬¤੢ Ԗ੢ զ‫ڏࢯؽ‬¤੢ 5‫׶‬ń [‘æ

mÃ˽̒
" ͕ * 4 ͕ ?͕ B ͕ ᙢᛦ
{
̩ĭ Ñ ȝ͕
Ă
#
%0bĔɅ T0°
‡0+Ʌ šc¿% Ʌ Ň
šŐɅ
-Ʌ 0&Ħ
Ʌ
Ʌ ['ijɅ ɅÃ"'ȳɅ

ɓƄ˾ ͕ V{͕ ƍ͕  ‹


ϩ 

ϩ
ŗ
tV ͕ V͕  ‹ 

ϩ 

ϩ
ɼ ͕ ˓V͕ əĂ ‹ Í

ϩ 
ϩ
ž£U͕ ͕  ‹ 

ϩ Ƀϩ

7VVü üVVü

z༐ ༐
ăŅ༐ 
 ‘
༐ e ༐ eÔ%Ȓ༐
ˊ
Ņ$ă༐ ǂ༐ %
༐ Ô%Ƃ

༐ e༐ !༐
%DB ༐ Ɣă༐ ༐ 
Q
༐ %༐ă༐ D༐ Ô
༐ D
ಎ<༐ ߻ѴDe
༐ QD༐ ༐
!%༐TL༐
༐e ă༐༐eÔă༐ Q†༐é༐%༐ Ņ༐༐
D >༐‫۾‬Ô༐

༐༐
%Ņ
༐ D
>༐Ô ༐D༐  ă>༐ƒ
>༐T
>༐ D༐D%༐
T
>༐ ă༐ !Ņ༐ 
Le༐ ༐ QD ༐ DÔ<༐ x༐
TăĐĜQL༐ ă
༐ ༐
ˊ
e ༐༐
༐ Dă༐༐ Q$༐ ăQQ
༐ ༐ ༐ .༐ D
༐ Ô"༐
%༐ ²
.ŵƹƣ༐D༐ ² Dă੍ًυϤ༐ƒ L༐.B༐༐  ༐ ༐e
༐Q
e e༐
!༐ %
҂༐ ıÔÚ‘ʺ༐ %༐ ëÔı  <Џɂɫ༐ Ø
༐ D༐ LඡੰÔ༐ ༐ Ŧɜ
}Ô"༐
ʺ! ҂༐!༐!TÔ༐ De༐D༐ eăÀ$҂DÔ
༐༐D༐
ɖɖ߼༐ Ô
༐† Lă༐ăe Ú
Ɗ ༐
༐Ņ
༐ ԙҳ‫>ݿ‬༐ ƒÓ༐ 9๑༐H†༐e  ༐ӚŠŠƊb˿੢ ༐҇Ɨ
༐ H༐༐ 6 ñĘ Šࢌb˾੢ ༐
༐ D༐ Ȁ>༐ Ô ༐ ༐ ɒ༐ăT>༐ !e༐ !%Ô༐ īĪ¡¡Ɗb ò੢ Ø
༐ D༐ LQÔ༐ " ă༐
EEǯ P 
>zǯ bǯ 
/
ǯ .>ǯ
z
>ǯ "
ǯ

ѯѯۭŅေ

૚Ņေ

ő
Ꮜᛦ
͔
<ອ Ѭေ
಺ᛦ
ಅອ
ै੢
w
 ढອ
ອ
Ŕ
ອ

Ӳԓ
Ꮞᛦ ѬŅѬŅေ

HC,+O H ) ?DMO

‫ ۔‬ŅŅေ ͶŅŅေ ϺŅŅေ ѧŅŅေ ͹ŅŅေ ͺŅŅေ

 7 ) C C O &,:>,?)DH@,O ࠥᛦ౅ᛦ

ǁ*pRΝ
ေ û n0Ò[͕ ΃55(Ŷ<*(ေ (nj

țေ ေ(
(
ေ p

ေ ေ íေ 5p
([(
ေ
Äေ
ေ ę8w(È
Ëwေ ေ Ą


‫ح‬RΝ
gေ Ň-
ţေ 4
w ò ɾ ‫ ڴ‬ေ¾ Ǻ Œ͕ †"Cǯ %4æ88*ǯ L‚ǯ ö) 88ǯ
*3*L*3ǯ
 *#_l¼0ǯ _Cǯ
ǯ"ǯ *#ňǯÓ%CL3'ǯ ÈÐÌ¥ǯ †£*#ǯ
#Nǯ ×i)4²ǯ;;ǯ "lǯ ‚% *Cdǯ
ª*i  N ǯ


༐ Çē༐
༐   ༐
D
 ༐ £

Ô

༐ ༐ 


༐ D ༐ 

༐ . †քĜ༐

༐

 ༐ £༐ ༐༐ ठ
༐
༐ 
<༐ Ø †༐

༐ Ô
༐༐



༐ %

Q
༐ 
༐ £ ༐ 
%
༐ 


༐ 
༐
é
Q
༐ 
Q
༐ £༐ £
༐
ą.
༐ 

D£9

༐ .

༐ 
༐ ƍ
༐ 
B༐ ༐ 
༐ 
༐
QÔ£༐ Ô
ɋ༐ ༐
 ༐ .†༐ Q༐ £Ô
༐ ༐ é9

༐ ༐ 
༐
Ô
༐ ༐ 


༐ D

༐
.
Đչ༐ £Ó†༐ £††£Ô
༐  †
༐ £
༐ £
 ༐ £༐ ༐ 
༐ Dщ
<༐ ϠD ‚
9τ ೪ Ò༐ ů±
༐ ె༠ትጏᛦðÔ(༐ ༐.D†༐Q༐.9༐%
༐ɣщ༐
͐ʀ
Q
༐
Q
༐
༐
༐ ംщ༐.༐%༐
 ༐ †༐†Q

༐ĐĜ༐
£†щκ༐ ƍ
༐༐༐  ༐

×
Ô
>༐ ༐
༐༐༐.
༐

Ô
<༐
 )ǯ ˋ1ԓ

ćĎ
ণࠣᛦ
੃ࠤᛦ
ࣹ০ࠤࠤ

ࣹࠤ Āᛦ ǶȂ
ࣹࠣᛦ

/

ࠣᛦ

ࣹ੄ᛦ

੄ᛦ

ণࠣᛦ

ࣹࠣᛦ

ੂᛦ

,79:7;6<>;7@ ేᏑᛦ
Ꮛᛦ

m*p8
ေ í g ေ ΃


(
ေ ေ 

(ൻေ ( Mေ ›Yေ ေ ‰5
<8ཀ
ေ ေ vĹRေ

Ƴ8
(
ေ A¿
ေѹ8ʡ‫ ڐ‬š ‫ھ‬Ӟൖေ …œ&+8ĕĖǯ Ä«œ7$œ8ǯ%&jęŻì7 j0ǯ";c* V ĉ ǯ

^=‫ أ‬ŕ Ŷ༐ = ༐
༐ X= ༐ ƛ༐ ’=B༐ ༐ 
L༐ 
༐ =9༐༐ 

༐ =༐
Q
 =༐

༐=༐$9༐
༐=9  ༐
!
༐=༐ !༐=༐༐=
༐
=༐ %<༐ ¯
༐ =
༐  ༐ .
༐ Q$=༐ =9
༐ %༐ ༐ £ L༐ Ƃ=༐
=$=
༐ =༐ 
༐ ༐


༐ 9 ༐%B
=9<༐^
༐!༐%9༐=༐ ༐༐

Q
༐ =.༐ 
$
9
༐ ༐ 9
༐ =༐ ༐ 
༐ =༐ 
༐ =༐ ƒÓ
༐ 9!
༐
 ༐
༐Q99f༐^ ༐9
༐
༐༐%
༐
ÇȔ
Q
༐
=Q
༐%༐.  ༐

༐ ༐
Q
=$
༐༐༐ ͖ ࢎ$

༐ =9=༐=༐ =ʄ9=༐  ï༐ =
༐ 
r

༐ 9 ༐ !༐ 
༐ .
༐ 9
༐  =༐ ༐ 
 ༐ =9=>༐ =༐ =[
) ťī 
6Rǯ R :ǯ 
/
ǯ .6ǯ

^ǯ  Ţš
ǯ

Dŗ

[Dŗ
ō֦

ࣥອ

সອ
.ອ
सອ
ϟອ
ˣອ
†ອ
‫ھڽ‬ອ

ᄞᛦ
ؓອ
Ǟອ
े੢
ᙧᛦ
࠱ᛦ
ढ़ᛦ
Œອ
߫߬ອ
ຄອ
ଙອ
ຂອ
†ອ
² Gˎ Gˏ Үү੢
ːGG²ˑ ˒

ল Ћ ЌЍȓᛦ

ೠ ᔦᒋᆿᅺᛦ (ƒ ጗ቼᅓჭᛦ‫ټ‬Т

¸[Dŗ +Iª«

 ŗ [ŗ ŗ ŗ ŗ ŗ

ဂᆿፖᏴᆿᐸ ჭᓅᔦᐹᆿᛦ V6ŗ
mp?<
ေ  û?ϩ ɤ:
ေ 
ǯ-ေ ęေ ǯ
ǯ
ေ p

ေ
ေ œԱ 
p
ǯƄǯ5ေ
ေ
ေ Rǯေ ęေ

ໝ
<
RΝ
g ေŇęŶ
<ေ4
Ƿgɪ ࢊေ ƞ ೝೞ ˁ ອ r?œ4Uà
ıǯ ?Ljǯ ©

ǯ Ý?Ll
1 Q ǯ

 `༐ >༐ 99༐ ď༐ `>༐ `9༐ `T
>༐ ༐ 9
༐   ༐  Ɠȣԓ
`༐`9³`>༐༐`༐
`Q
༐ ༐?༐9
༐
༐`༐ 
<༐Ĭ༐
]
༐.?
༐ .  ༐ é9
༐ ?`༐ †༐ 9ā
༐༐ 
༐ 

9
༐†
༐
༐ `9༐```༐ `༐ 
༐d9†༐``༐`༐
༐%!B
`9༐
>༐ ༐L༐
`?༐ `%
༐ Q
༐ 
༐ ?`༐ Q`9L༐ `༐ ༐ 

L༐  ༐ 
༐ d9
༐
``༐ `༐ 
<༐ ]

༐ 
༐ `༐ .༐ Q†༐ ༐ L ༐ `LƄ༐ Ǵϖອ `༐ ‚

 ༐ ?

9
༐ ༐ ?
༐ LL༐ . ༐  †༐ `
༐ .Ɣ༐ %
༐ ༐



༐ %༐ L༐ ༐ 


༐ `༐ 
༐ 9
༐ 
༐ †<༐ ]
༐ Q9
༐
æ +
:ƛ
Ƅ ɑɑ ພေ ေ
ေɊ ေë ÄÀ{Rေ +gေ
ေ Ƿ
Äǯ<`
Ƿေ
ေ pƛ
ေ ǯ

pေ
yR-ေ: ǯ ေေ?ǯ ေ
ေ
Ɏ
ධËv
ေɏ
ေ
ေǯv
-
ေǯ ǯေ
ǯǷေÄRŶÒေ

ေǯÄေÄေÚՐေ Á ɯ š ͕ Րေ{Xȍʿေó ïϩ ‫


<ܗ‬ေ ‫ٶ‬Ʋ

yj  ေ 9ƶʣ͕ ư

<gေƲɏ
ေÄ?È
¿ေေေ`
ေR
ǷေǯƛǷေ
Ƿေ ေǯSÄ-ûေǯ
ေ`5ေv
<5Ƿေv
ေ:Ƿ
ေĞŶ g ေë 5ေ5Ä5ေ

? ေ
ေRÄ
ǷေɏÄေÄRေ-ေ
Ƿv
ƭp
ေေǯ

pေp
ÄÄေǷÄေ<ေေ-
ေǷ
SÄÈ
ေęေေǺǷ
ေ‫ތ‬ေ
ǷေwŶ
Ƿေ
ေ
Q
ǷေS<v

ေေ
ေy?
ေေ-
ေȃÄgေ
 )¦ǯ )

ອ
a

%੢

7੢

˷ͤԓ

ϙϚ֦

੨ေ
ອ
কອ
DE   a
х֦
ϛϜ֦
཭ᛦ
ᙪᛦ
ľອ
বອ
%੢ ͖

7੢

a

a  a
a  

P8NY8J6]a ^6A 8 S a‫ޤ‬ᛦ ,86

a

ǁpR
ေ - $ T@> % x :ƚေę

(ေ
ေ ေ 
S

खR
ေv
‫׬‬ԙԠ‫׬‬ေ ęc
ƳR
wě[ေ ęေ ‫׽‬ƌ(
ေ
p
ЋǶɃေAęߣ
cေ4
k Ė džਖ਼ေ ¾ Ñ þ͕ rM&+
Cǯ&3ǯ );

ǯ Ý&l
1Nǯ

q༐
༐ é9
༐ 
༐ q༐ .
༐ ༐  †༐ 
༐ 

༐ q༐
%B†9༐ ༐ %
༐ 9
W༐
༐  ༐ ༐ 
༐ 

%
༐ ༐ %༐ 
༐ 9༐ q༐

.†༐  ༐ Έ༐ÇȔ9

༐9༐ . ༐ 
༐ ༐ ༐

9
༐q
༐

༐%‚
༐ %B
9<༐ z༐ 
༐ q9
༐ %

༐
.༐ %༐ Ø༐  ༐ 
༐ é9
༐ q༐
.
༐q༐ 

༐ 
༐ ༐ 
Q

%
༐

<༐ zq༐
༐༐ †

9
༐ q༐.
༐༐

༐
99 ༐

༐ 
༐ 

༐  ༐ ༐ Q†༐
d9%9༐ Ѷ9
>༐ .
༐
ÇȔ9


༐༐
༐ 

ū༐Eq༐
༐༐

9ā
༐
༐

>༐ .†༐ȯq9


༐ 9<༐ Ø
༐

༐༐.  ༐
༐ÇȔ9
༐9
༐
†
༐༐ †༐༐

9
༐ϱʀ


༐
༐ 
༐ 9༐ 
༐ 

9
<༐ ¶༐ 

 ༐ ༐ 

༐
Q

༐ ༐ ༐ ༐
˄ǚ ††>༐ %
%
༐ .
༐ ༐ %
༐ Q
༐ 9༐ ༐ 

༐ 

9
ij༐
༐
ā༐ Q99༐ %B
9༐ .༐ 
༐
ÇȔ
EQ
༐ 
Q
༐ 
༐ .
༐  ༐ 
༐
%

༐ %
%
༐ ༐ 
༐
9q
༐ 9 ༐ 
༐
༐ 
༐  ༐ q̤.
༐ v †༐ Ѫ༐
%B
9༐%
q
༐
༐9%
༐
༐

༐༐
༐9<༐
gǯ P 
>h÷ãǯ  ƂĊǯ 
‘/
ǯ .>ǯ

>ǯ "
ǯ

)

)
)

)

ᙟᛦ

)
̤ԓ #)
͸Γ  ) < ''[

ʾԓ  )
໤ᛦ
᎔ᛦ
ᕔᛦ
ᔫᔬᛦ 
 ,
%)   )
)

)

 )
   ,''[

) )
) 
)
)  )
೟ྑೀཟိ ೀ໼ၻ С ొ཈ྻ‫ٹ‬ᛦ
ః‘pR¦5ေ  ?ϩ ġs<s‰‘ေ ęေ ư
ေ s(ေ Ôðgm g Ē M ေ ǂʆππေ ęs(Ë<ေ Ɔ ࣸ ʺ ٥ २ ٦ Ǡ ᛦ swေ Ƿ‘55(вေ•‘(ေ
(w͡ĥsေ ‫›ٷ‬Ēေ ư‘‫ ح‬ေ 
ƏR
Ż gദေ ¬ေ `Ћ‘‘(sË
ေps//ေ ɫ+Я‘wpေ Xေ IIĒ qေ ƚsေ pƭƴƴေ
औ+Я‘wpေৈ ࠠ ٣ ᛦ Ē Mေ swေ
ေpsƴƴေ Ô+ϛpေˏgေhҿ zေ Ň-
ေŠ8[
< g ࢋ ࿵ ေ 3 .?æ ąèU
Cǯ
ǯ*Ǚ++MUǯƧǯ> #ǯ
*3*8  Q ǯ

G

ü ü

Ø †༐ 
 d9†༐ `༐ `U ༐ Qජ`9༐ ༐ ?
༐ `༐ `.

༐ W༐
`༐ !༐>༐ ༐9
༐ `༐`༐ `༐ 
༐`9`༐
༐ %9༐
Q
༐.
༐ ༐ 
༐
%``༐ .
༐ 
༐ ༐  
༐ 
༐ 
d9
༐ ༐ ༐ d9!
<༐
ƛǨ༐

༐.? ༐༐9%
༐9%
༐`༐%

༐
༐.
༐
༐
dR
༐
!
༐༐ ༐!
<༐ Ø
༐`༐༐%
༐
༐`༐  
༐༐
†༐%༐
9`>༐ `༐ `༐ 9%༐ †††༐ 
༐ %`
༐ !U༐ 
༐  `9 ༐
༐ ``†ȧ` 
༐ 9!
<༐ x༐ 9%†༐ `༐ 
9`༐ 9༐ %†༐ `%‫ޗ‬

Q†༐༐9
9Ǩ༐ `9༐!
༐`༐
9Ȁ<༐
Ø †༐ ບ`༐ `༐ !
༐ ༐ %9%%
༐ ഒ༐ %
༐ †9༐ !Q`
༐ %
!9
༐

Ǩ
༐ «
H༐ ?
U༐ `༐ `

༐ ༐ ?
>༐ 9U ༐`‘`༐ ༐ %
B×
`. ij༐ Ġ
༐ 9`༐ ?Ù%9`ĐĜ༐ `༐ ༐ ! †༐ 9%
༐ `༐ Q
༐ !Ǩ༐ %9%%†༐
Á
Q
!༐? `9ҳ༐ 
༐9%༐ H†?Ï© Ƅ༐††ƥ༐`༐9 ༐
<༐
ѳ Ï༐ `Ï˾ఖ†Ɠ೫༐ `༐
T!‫`݌‬U༐ `༐ ? †༐ †ଳ༐ `༐ .  ༐ 
༐ `༐ ༐
Ï༐ Ҭ `9༐ ሷᆸᛦ HÏ༐  U༐ Ġ༐ `༐ 
༐ HĖǼ༐ `༐ 
༐
Ė༐ †`QH༐ `༐
Ġ†༐ H༐ `9<༐ x༐ `H༐ 9%E༐ ༐ †d9†༐ 9
>༐ <
<>༐ `%
U9༐
 ""O QҶ੢

֣ƨ̤Ջ
>༐ !༐ ࡘ9ǰ
༐ ༐ !༐ Þ! £L>༐  Þ ༐!༐ !£ ༐ ༐ \ÞɄ༐ ÞL<༐
x༐ T9Þ
༐ ༐ !༐ 9
‫݌‬༐ ༐ B!༐ ༐ .!
Þ༐ !༐ Þ! L༐ !༐ !༐ %LL༐
9 L
 ij༐༐ 9༐ !Q
༐!༐9!%
༐!
Þ༐༐!!L༐ÞLQ!༐༐ L༐
༐
!༐!Þ
<༐ƛÞ! 
༐༐ !Q
༐
༐!Q!! L༐༐T‘ ༐༐Þ!Þ༐
!
Þ
>༐ !༐  9༐ !Q
༐ !༐  Þ༐ L༐ 9
༐ !༐ T
[Þ

༐ !
Þ
༐
༐ ÞQ
<༐ ’
༐ 
!༐ 
!༐ !Þ
༐ ༐ %
༐ 
༐ ༐ £
༐ !༐
!
Þ༐ !༐

༐!
༐
>༐£

༐
!༐!Þ
༐ÞL
Þ!%༐Þ
%
!
 Ƅ༐Þ༐!༐ȇ
!༐Þ
T‘
<༐
’
༐ !༐ 

Ȝ Þ9 ༐ ! LÞL༐ 9Þ!L༐ ༐ 9
༐ Þ༐ 
༐ Þ9༐ ༐

༐
!
Þ༐ ༐ !༐ 99༐ 
9
>༐ 
༐ ม༐ ༐ ƒ
༐ !×

Þ
༐ !༐ ༐ 
.༐ 9༐ %
༐ !Þ
9༐ Þ
<༐ ૪!9Þ!༐ !༐ !༐ %
༐
T
༐ . ༐ !Þ༐ ༐ !༐ Þ!༐ ༐ ş <7 ʹ ༐Ū iƪ༐ ༐ %£!༐
T
Þ༐ !༐ %༐ !Þ!༐
৻%9 ƴ༐ 
༐ !ÞT!L༐ ª༐ ༐ £ ༐ !ǰ༐¾༐ @@ XÛû >༐ ƪ༐@ XƋ>༐
™༐@  5 >༐ 
༐Þ
!
Þ༐ %L ༐ Þ
<ɂɂź༐]
༐

Þ!9ā
༐ ༐
༐9Ɇ!
>༐
Þ༐ !༐༐. ɘ ༐
༐༐༐Þ!
>༐ ༐༐
༐Þ
Þ༐༐DŽèèèºX<༐
]
༐ !༐.
Þ>༐.  ༐༐ÞL!Þ
༐ %༐ Þ9  ༐ !༐9
>༐Ĭ༐

Þ!༐
!༐ !༐ !Þ
༐ ‘
༐
d9Q!
༐ ༐ Ŕ Ǖԓ Þ༐ ʯ ώ7༐ 
 ď༐ i
<༐ 9 ༐  !༐ ༐ .༐
༐ !༐  Þ9 ༐ !༐ ˜7ä Þ༐ ƟIJ7[
 ༐ Þ

<༐ z༐ ༐ Q
Þ༐ Þ!༐ ༐ 
!༐

༐.
Þ༐!Þ
9̤༐ ༐ !֪༐ !!༐ .  ༐ .9༐ !ƨǚ
༐
T
QL༐
%9%%
༐!༐!༐ 
Þ༐
༐
Þ ༐Þ
<༐’!  ༐༐9
༐ Þʢ9Þ༐
!>༐.
༐%༐Þ
༐༐.!
Þ༐!༐!

>༐Þ༐! >༐!༐ [!Þ༐Þ >༐
༐ 
༐ !

༐ Þ
9Þ
<༐ ] L༐ 
!༐ !༐ .LÞ༐  9༐ ༐ %
༐ £ā
༐
Þ
༐  !༐ 
༐ Þ༐.༐ !༐ %
ÞL༐ 9
<༐̄༐ ²!Þ༐ !Þ
༐  9༐ %
༐ !ÉL༐ ༐

༐ 
! ༐ ༐ ƒÓ
༐L!༐ ̤L!ǰď༐ !༐ ƒÓ
༐ !༐ ǰ 9ȇ༐ ŦL༐ 9
༐ ǁ­­ࠂອ !£hÞ༐
Ê
! g༐
’
༐ !༐ ༐ ÞɄ9༐ ༐ !༐ QLÞ༐ ˾ഘ
༐ .LÞ༐ !༐ !༐ ǰ!༐ !9༐ ༐
d9༐ %LÞ༐ !
>༐ ༐ !༐ %L༐ ÞөÞLL༐ ༐ Q!Þ9¾༐  !
༐ !༐ 
ê༐

Þ
༐ !༐ !༐ 

Þ!9Þ
༐ ȇ༐ !%Q
༐ 
༐ £
 ༐༐ ༐ 
༐ !<༐
x༐Q£Þ
9༐%༐Þ
9ƒ༐.  ༐ !༐Þ9 ༐
༐!
༐Þ
Þ£
༐!༐
༐!Þ
༐
!>༐
T
༐ !༐༐.༐%
༐£Þ!9
༐Þ!
Þ༐ !༐Þ!!ÞL༐%L!9
༐༐

༐༐ Q༐ .  ༐ Þ
!<༐ ΀!Þ
Þ༐!Þ
༐ !༐ !༐ %
༐Þ9
༐%༐ ͞
! >༐!Þ༐༐
༐
 ༐9
༐༐
༐Þ ༐༐LÞ!༐.!Þ
༐ ÁX !×

Þ༐  ñ ƕȪƔǸȴԓ ԥ.
Þ
λ !༐ 
༐ !Þ
༐ !༐ %L༐!L༐༐ !༐ 9 ༐ 
Þ༐
‘L༐ 
Þ!
༐ Á Ɉ ƪ ༐ ©༐  !༐ ༐ !!!%L༐ Þ༐ Q
!༐ !.!Þ
༐
.
Þ
༐ Ɉ ‫ز‬༉‫ث‬༐<༐ ༐!༐Þ
!༐
Þ!
<ɂɂƣ༐
ƛ!༐.
Þ༐Bď༐ !%9£༐
[ !༐<༐  B༐!༐ ^ԓ <༐!LLÞ༐!Þ
༐9
༐
Þ༐ 
༐ 9%Þ!༐ ༐ 

༐ 9Þ ༐
T£Þ9<༐ ] 9༐ !༐ ƪ}<༐ !

Þ༐ £

༐
%!Þ༐!༐ %
༐
TÞ9
༐  Þ9 ༐ !༐ !
Þ
༐ 
༐ !༐

Þ L༐ . ༐ !༐ !ƂÞ༐ Ȝ
Ǻᛦ Š
Ћ/ေ (8
ေ /ေ
ေ‰
ေR/
ေ ေ`Q
%ေ
u(
//%ေေę
8[ေ K
//Mေ ư ( ေ/ေ

‰
ေ
ေ
ေ/8`/
Ƴ85ေę8
(
ေRေ
̯ေ|
ေ


-ေęေK
//ေư
ေ/

/ေေ‰ 5ေ

/R¦

ေ ęေ
//-
k(
ေေ
(ɏ
(
ေ
‰
(Qေ/ေ
/(`
ေ8
ေ
‰ ေ Ᏻ Ꮘˍýᛦ ေA4|౯ေ4¿

ေಢ
 ေေ|
/ေ
+ေ——À2Mေ
ေ/
v
/ေေ

/Ā5ေ
ေ5
ေေK

ေ(

ေေ
KေKƛ
//ေ:5†ေ
Ē˜༐ PH6Rǯ  :ǯ H/ Hǯ .6ǯ HH6ǯ "Hǯ


༐ ༐ Ɨ†༐ ༐ %9#༐ y༐  <༐ QǦ %༐ %9#༐ ˜7༐ †£
༐ ò@ ༐ ༐
.
  ༐ y77༐B ༐༐LT9
༐༐£.༐༐ 

༐
 ï༐9 ༐ ༐
>༐
%9#༐ ş77༐ ༐༐ ò༐ ༐ 
ò#
༐ ༐ ༐ ༐ ˜7ǟ@Ɨ B༐ ϴò༐ ༐ 
༐

T9L༐%<‫ٱ‬ʴƹ༐

üü
]
༐
£ ༐༐Ƃ @ ༐#.༐

༐༐
ò༐%༐༐ò.}
ò ༐òò>༐B.༐
༐ò
>༐ !༐%

༐B.༐༐༐
9
>༐%9༐£
༐
 d9
༐༐Ƃ @ ༐
#.༐

༐ ༐ ò༐%༐ ༐ ò.Ɨ
ò ༐ ò!༐ ༐%L
༐ 
:Lò
༐ @ò༐ 9ā ༐

༐ ༐ †ª
༐ >༐ 9ò༐
ˡ
ò>༐ !༐9༐:
༐ò#
༐ʠ
‫ڼ‬ò༐×
òªg༐ Ϝò!༐ ò
>༐ ༐ >༐ ༐ ༐ ༐ òò
>༐ ༐ %
༐ !ò
༐ £༐ ò༐ #༐
#@ ༐ ò༐9L>༐ . L
༐ ༐9ª
༐ ༐ :99ȧ ༐ 
ò>༐ %9༐ ò༐ #༐
Ƃ #༐%
.
†༐ ò༐@༐
#ò༐༐%
.

༐ʠ༐༐ ò<༐
Û
༐༐#
༐Ç#༐༐
ò༐
%
༐%༐ë@‫ڌ‬ʴŕ༐༐ Ǐ •‡˜༐9
༐༐ò.lj
ò@ ༐
ò!༐9£%ò
༐༐#
༐— @ ༐༐༐#༐ ò༐༐
!ò༐༐༐ £

#9
༐༐
Ǣ77ĵX < ༐ ] ༐ 
 d9
༐ .༐ ò òò༐ 
:
ò
༐ #༐ 
༐ 
ò ༐ ༐ ò!༐ ༐
ò༐ %༐ # L༐ .B
>༐

òò༐ ༐ ü9
>༐ ༐ ༐ 

༐ 
༐ .B
༐
༐
༐Ʀ<༐²<༐ :
༐ò
༐

ò:
༐#༐#
༐

༐༐# ༐
 d9
<ŵʴʴǕŕ Šʴ༐
˛#༐ ༐ 
༐ ò
༐ ò
༐ :
༐ ༐ ò
༐ TL༐ £
༐   ༐ ༐ Ē7༐ #༐
Ūƪ༐
༐ 
>༐ %༐ T
༐ ༐ ˜7༐ ༐ ‫ا‬༐ 
༐ 
༐ ༐ ò༐T
༐༐ y7༐
£༐ƪ༐ 
༐†£>༐༐ 
༐£բ
ò༐
T༐ 
ǃ
£༐༐
༐༐ Ū7༐£༐
ş˜7༐ 0A Ǐ 7‫׷‬ɹ༐੎X<༐ȭ
d9
ò>༐༐òò༐༐༐‘༐T
༐ ༐ ò99༐Tȸ
†༐༐!ò༐9 †༐༐༐
༐

༐
£
༐༐#༐:
༐#
༐
ò༐
#%ò༐༐ £
༐ ò
j༐ ]
༐
T!༐
ϴΓ
༐༐ £
༐ ò
༐ ò!༐9#༐

†ò༐ %
༐ 
༐ £༐ 
༐
T༐ 
ǃ
༐ ༐ £
༐ L༐ ò
>༐
%9#༐༐

òò༐ 
༐ £༐ò
༐
. #༐ %
ò.༐ 
༐
T༐ 
ǃ
༐ ༐

༐ ò༐#༐%
༐Ƃ 
 <ɂŠŕ༐
΁L༐ ༐ 
༐ ò༐ :
༐ ༐ ò
༐ ò
༐ ༐ #
༐#༐ ༐ 
ò༐
༐%
༐
༐༐༐9%òò༐ò.
༐#

9
>༐%
.

༐‡77ĵ༐༐ƪ77ĵX>༐
 ༐༐˄ǚ%ò
༐.£ ༐:
#ò༐Ć
Ɨ
ò ༐
 d9
༐.

༐#

‚
#9
༐༐Ū77ĵ༐£༐ş777ĵX༐
༐
d9
<༐ ²
ò༐༐£ 9༐%
༐
༐@༐༐ :
>༐
.  ༐ %:#L༐ 
༐ 9ă༐ ༐ Ć
༐ ༐ #
༐ 
༐ ༐ : ༐ %Ƃ
%ò
༐

༐

༐#
༐ 9%
<༐ ]
༐ò.
༐#
L9
༐ ò༐‘
༐ £ L༐
༐ ༐
T༐ ༐  L༐ 
%ò<༐ ˛#9
༐ 9%
༐ ༐ 9%9
༐ 9%
>༐ ò༐

 }༐ #9%
ď༐ .  ༐ L༐ £˗ϩòò༐ 
#:
༐ #༐ 

༐
ÇȔ
>༐ 
༐

༐.
L༐ò
༐ ò༐ 
ò ༐༐%
༐ɭǚL༐༐ 
#༐:
<༐ ]
༐ £
 ×
d9
༐ ༐ò༐%

༐ò
༐ ༐ ò
༐9%
 ï༐Œ 9
༐ y <ƟIJƪ༐ .༐༐Ć༐£9
༐
#9%
༐ .  ༐ ༐ %

༐ 
ò
༐ %༐ # ༐ 
# <༐ ড়೬༐ ..༐ ༐ 9£
༐
9%
༐
༐#
༐#༐ L༐
:
ò
༐༐ ༐
>༐༐

༐
:9ò<༐
x£
༐
༐9
#༐༐
༐ò
༐ ò༐ :
༐%

༐L
༐༐༐ò#9༐
༐ ò:
༐ 9%ò
༐ #༐ ༐ #

£9
༐ ༐ %9༐ y777ĵX>༐ 
༐ 
9ò# ༐ ò
༐
ò༐ ༐ 9
༐ #༐ .
༐ ༐ 9༐ ༐ .
༐ ༐ %9༐ Ay77༐ 
 <༐
 @]Ƀ ̀Ƀ

æ]k̚ &*- à XĀ̚ h$+˫̚ ±+ɟá̚ >¹Ô+İʺ˜ÝŌ Íċı!Ō ´ČóþĻŌ Cŗ ¿Ę!ôĽĩĦčļIJŌ )‰k·
°F
¶IJ}Ƀ

$C5„ c`|35h„ Dk„ ulu,UP€„ ]D~53„ |DmC„ ,„ *`P,mEP5„ PDduD3„ v^3„ 5DmC5g„ lch,€53„ ʐʠĦϰ
@`|53„ `^m`„ mC5„ c,hm„ m`„ /5„ l5,P53„ D^3D^B„ ,B5^mk„ ku1C„ ,l„ ^Dmh`15PPuP`l5„
,h5„ l`]5mD]5l„ uk53„ m`„ BDz5„ lmg5^BmC„ m`„ mC5„ c`|35h53„ BP,kk„ P,€5h„ /um„ mCDl„
^515kkDm,m5k„]`h5„1,h57uP„:JhD^B„m`„ 5PD]D^,m5„ nC5„`hB,^D1„],mm5h „%C5„c`|35h„
],€„ ,Ukb„ /5„ ch57`h]53„ D^m`„ ,„ kc51D,P„ kC,c5„ kw1C„ ,l„ ,„ |,kC5h„ /5,3„ /,h„
`h„ 3Dl1„ ,^3„ PDBCmU€„ <Dh53„ m`„ BDz5„ Dm„ lmh5^BmC„ ^5„ `7„ mC5„ /5lm„ ]5mC`3l„ Dl„ m`„
A`|„ mC5„ PDduD3„ BP,kk„ `^m`„ mC5„l2,PD^B„luh7,15„,m„mC5„ mD]5„ mC5„,hmF1P5„Dl„ 7,/hD
1,m53„ ^„ mCDk„ |,€„ mC5„ ,hmD1P5„ 1`]5l„ ,Ph5,3€„ ch`zG353„ |DmC„ ΂ϰ Q,€5h„`7„l`Q35h„
BP,ll„ D^„ nC5„ hDBCm„ cP,15„ ,^3„ h5euDh5l„ `^P€„ C5,mD^B„ H^„ v^„ `z5^„ m`„ 1`]cP5m5„
(Z„
mC5„ l5,P 5^5j,PP€„ oC5„ c`|35h53„ BR,ll„ 1`,mD^B„ ,ccRI53„ m`„ mC5„ ],mD^B„
")[7,15k„ Jk„ duDm5„ mCD^„ ,/`um„ ­̄ ]DPl„ mCD1O„ k`„ mC,m„ mC5„ luh7,15l„ m`„ /5„ l5,P53„
]ulm„ /5„ `hDBD^,PP€„ ch5c,h53„ m`„ ,„ 1P`l5h„ m`P5h,^15„ mC,^„ Dk„ 1ulm`],h€„ `h„
^515ll,h€„8`h„c,hml„N`D^53„ /€„=\v]5„k5,SJ^B „&CDk„h5fuDg5]5^m„],O5l„lbP35h„
BP,lk„l5,Pl„ ]`h5„ 1`lmT€„
,PPuc„ ,^3„ D^B|,P+„ C,z5„ J^z5kmDB,m53„ ,„ ^u]05h„ `7„ k`P35h„ BP,kk„ 1`]‚
c`kDmD`^k„ kuDm,0P5„ 7`h„ 5T51mh`^„ mx/5„ -ccUD1,qD`^k„ ^3„ 35m6h]K^53„ zDk1`kDm€„
mC5h],P„ 5c,^lD`^„ 35^lJn€ „ ,^3„ D^35„ `7„ h57h,1nD`^„$yP5k„ ÕŗxWŌ \ ^ \ \ x ̄ǑÆ\M̄
,^3„ \^\…̄ ,h5„ m,O5^„ 9g`]„ mC5Dh„ h5c`hm „

P,kk5k„ „ €ÀɃ ,^3„ z̄ D^„%v/P5„ \ ^ \ \ ̄C,{5„/55^„ ul53„ ,l„sC5„ 15]5^mD^B„]53L.„
D^„ c5h],^5^m„ ],hOD^B„ D^Ok„7`h„ km,]cD^B„ m€c5„^u]/5hl„ `^„ 3D:;5h5^m„ OM^3k„
`7„ 5P51mi`^„ mu/5l„ |CJ1C„ uk53„ k`7m
BT,kl„ 5^z5P`c5l„ C,zD^B„ mC5h],P„ 5c,^ƒ
lD`^„ 1`5>1D5^ml„ `7„ , ![J],m5P€„ ©Ō ÜŌ \ņīǥdž¢Û̄ 51,ul5„ ^`„3D?1uPm€„},l„
5c5hD5^153„ |DmC„ mC5k5„ nCh55„ BP,ll5k„ Dm„ ,cc5,gl„ mC,m„ 1`5>1D5^m„ `7„ tC5i],P„
5c,^kD`^„ Dl„ ^`m„ ,„ 1hDmD1_P„ 7,1paV„ D^„ mC5„ ul5„ `7„ l`T35V„ BQ,ll5l„ DVW„ ]'VODVXB„
M^Ol„ ,ccPD54„ D^„ mCD^„ Y,€5V#„ `{5h„ PD]Dm53„ ,h2,l„

P,ll5l„ Ō ,^3„ ÍɃ C,z5„ /55^„ ul53„ D^„ rC5„ l5,PD^B„ `7„ P5,3BP,ll„ lm5]l„ m`„
Ɏ‡༐ P 
ǯ ūƟ:ǯ 
/
ǯ . ǯ

 ǯ  "
Ŵǯ

4 ϩ è  Žϩ = ċ ϩ ’ ϩ ”Ýϩ c̐Aϩ ;ϩ ¿G*¯ 7ϩ

ċC–Ʌ ’Ŧ öɅ R–ŠɅ
i Ʌ +
ĩɅ ]=Ʌ CJ]+Ʌ +
ȟ ½Ʌ ]>Ʌ CJĒ+Ʌ n
ĩɅ ]=Ʌ CJ]+Ʌ
}&ˆɅ ] ÓɅ #&Ʌ ] ÓɅ &Ʌ ] ÓɅ

+Ĩ  ' ϩ +@ϩ / 
ϩ /  + ϩ
ϩ /  / ϩ /'  + ϩ @ϩ ϩ
ϩ +
  ϩ +
  ϩ   + ϩ   / ϩ
'ϩ /  / ϩ /  + ϩ +   ϩ +Nϩ
ϩ //   ϩ // 
ϩ @+ϩ  
ϩ

ϩ + ϩ2ϩ + 
ϩ ϩ @'
ϩ   2 ϩ @
ϩ
+ϩ /   ϩ / 
ϩ +ϩ
7ϩ AęË
ေŠ
8ေ
ေĸ:
eً Ŵĭေ

4 ŭĉϩ è    ϩ lA;ϩ `ϩ c¾¯ϩ s ϩ = ċ ϩ … 7ϩ

 0(
6< #$< $;'7< < < < !<
R zJ
Ʌ B
Ʌ R±Ʌ
 ?Ʌ *
ȜɅ 
JɅ    Ʌ
 JɅ  Ʌ A Ʌ
Uª7Ʌ Uª7Ʌ Uª7Ʌ UĉsɅ UĊ7Ʌ UĊsɅ UƺsɅ

'
ϩ ɤ˲Ȧèေ

ϩ 4ȍĭေ 
/ϩ  
ϩ Ĩ   ϩ    ϩ  
ϩ 2
  ϩ '+ 
ϩ

ϩ Ƽ`‹ ေ 
ϩ 4ȍĭေ '2ϩ +   ϩ ''   ϩ '

ϩ 2+ ™
ϩ /'   ϩ + 
ϩ

ϩ ɤ`‹ „ေ 
ϩ Vĭ‹уေ ϩ +
  ϩ 
   ϩ 2 
ϩ 
/   ϩ 2/ 
ϩ /2 
ϩ

ϩ Ƽ`‹ ေ '
ϩ śȿę ļϩ /ϩ + 
ϩ + 
ϩ 

/ 
ϩ +   ϩ '/ 
ϩ 2 N  ϩ

2 ϩ ɤ˲Ȧèေ 
ϩ Vĭ‹ϩေ 
ϩ /   ϩ +   ϩ + 
ϩ 
/ 
ϩ 


ϩ 2 
ϩ
 +ϩ ɤ˲‹ ေ 
ϩ ‫א‬ĭďуေ 
/ϩ 4ďĭေ
ů֦ '++ϩ '

  ϩ 2   ϩ '/   ϩ '' N
ϩ  N
ϩ è   ϩ


ϩ ɤ˲Ȧ èေ +N

ϩ Vĭȍʿ „ေ 
'ϩ 4ȍĭေ +ϩ @ 
ϩ 


ϩ  
ϩ '   ϩ    ϩ  
ϩ
ອ A
ေŠ
8ေ
ေ€:
† š Ŵĭေ
ɀေ ĮǺ 


ေေ

8
ေ;
8
 eေ

4 ϩ Ĩ  ϩ ”" Hϩ sϩ = < ϩ … ϩ Òϩ =Ċ Ċϩ

’ Ŵϩ ϩ 4* ˒ĉϩ lA;ϩ Ú^ 7ϩ
-Ʌ R ™J
ǶɅ  ǷJɅ B
#Ʌ JɅ R ±)Ʌ  Ʌ
UĉsɅ ǝ sɅ Uª7Ʌ

45ေŠ
Äေ ++ϩ 

  ϩ 2   ϩ
+ɲေ

/
ϩ ‫
מ‬ေ
2
ϩ 
ϩ '+/ϩ
+Яေ

ϩ ‫מ‬ƚ
ေ

ϩ ϩ 

ϩ
7ϩ A
ေŠ
ʼnRေ
ေ€:]eࠣ Ŵ ΢ေ

4 ϩ    N ϩ D ØǑϠ sa¡AçǒijΆǓϩ Daϩ 4çΚ ϡ¢ ϩ ʪ©Ɓ˻̙ Ęϩ ʇD ϩ

Öˀ΅*XÚŴċϩ ’ ĺϩ
  J Ʌ 
ǪǸ/
Ʌ zɅ ȧn
 #Ʌ
Įȕˆ) Ʌ
 5?Ʌ Uļ 7Ʌ †?+Ʌ < "8< J Ʌ "9,<

ñ ǭϩ '
ϩ ɤ`Þေ

ϩ 4‹ঝေ +  ț ϩ ɤ˲Ȧ èေ /+ϩ =¬ÜŁϩ +'ϩ Źေ 
! yϩ * eįೇ+ေ
c¯ϩ

ϩ Ţ`‹ေ 
ϩ=¬ęŀϩ /'+ϩ ɤ`‹ ေ ϩ 4ȍĭေ 
/ϩ Źေ 
! yϩ
c¯ϩ 
ϩ ɤ˲Ȧေ 
ϩ ‫א‬ĭ‹ϩေ +
ϩ ɤ`Ȧ èေ /ϩ Vĭȍ Ŵေ 2ϩ UǶ 
ȍyϩ
ñ'¯ϩ

ϩ Ţ`‹ေ 
ϩ ɊʤȦуေ / + ϩ Ƽ`‹ ေ +ϩ Vĭ‹ϩေ 
/ϩ Źေ 
! yϩ
cϩ
2Nϩ ɤ`Þေ 
ϩ ഏʸ༝ ʹᛦ //ϩ ɤ˲Ȧ  ေ ±  ϩVĭ‹уေ 'ϩ Źေ ±
ùyϩ ȉ[ေ ေ /
όǮϩ
c
Ĺϩ ' +ϩ Ţ`Ȧေ 
ϩ Vʤ‫ ٵ‬ը ұေ +ϩ Ţ`Þ  ေ 'ϩ ‫א‬ĭȍ ұေ 'ϩ =¬½ŀϩ /ϩ Źေ 
! őϩ

?/ϩ 4ȍĭေ +ϩ UǶ 
Ȏyϩ
Ǣ+ Ĺϩ
ϩ Ƽ`‹ေ +
ϩ Vĭՠթϩေ /ϩƼ`‹  ေ +ϩVĭՠթϩေ õ +   ϩ=¬½Łϩ

'ϩ 4‹ঞေ
7ϩ A"
ေŠ
8ေ
ေ€ :
 eš Ŵ ĭေ
 )Öǯ Ɏƪ ༐



༐


<༐ƛ

༐š༐Q
༐
༐ L
༐

༐ų༐ 
༐.B༐%
9

༐
༐E
༐

E
༐.B ༐
༐Áƪ77༐༐š77ºXģ <༐ ƛ

༐Ɏ༐Q

༐༐%L
༐
 L[
T
༐ ༐ų༐¾LLǨ༐ @>༐  LųLL>༐ !T“!༐ vă@
@༐ ¾³෰Ï
ǰ

༐
༐
.
༐.
༐༐
༐ ‫݋‬L<༐
Öƛ

༐ͣӗ༐ !
༐
༐%L
༐9
9༐༐
 ༐E༐༐¾༐ 

༐༐༐ ă‚

[೭༐ L༐ ˗ǂ¾<༐ Ԗ༐ 
༐


>༐ ?
ʔ
T༐  ༐ .
༐ 9@༐
༐%
9¾
༐༐ 
༐
༐
T
Ɇ༐

༐9


<༐ƛ

༐ 
༐

༐༐ 

༐ ͣ࠯༐ Q
༐%

༐ 9
L༐ ༐ 
༐

 ༐༐
} 

༐ ! 
༐༐
 
}༐ųE
<༐Œ༐


༐

¾༐¾
༐
T
༐

༐L༐


<༐
ċz༐ %E
>༐ 
༐
T

༐Q
༐ ༐༐ ༐


༐
T
༐ 
ă༐
༐.  ༐ Q
༐  ༐ Ǔ
༐ ༐

ų༐ 

༐Q
ų
༐
༐ 

༐
9 ༐ 
༐ 9༐
.E ༐BE ༐EB
>༐༐ 

༐༐ .  ༐ 
༐

[ 

༐

Ǩ༐.
༐B
༐9
༐
 ༐ų

E༐%༐༐LT
༐


༐

<༐
ÖĄ
༐ ༐ 
༐

༐ 

༐ Q

 !vL༐ ༐

༐ 
༐ 
 ༐


  ༐༐ L༐
E༐

%༐༐
༐
༐ Ǩ

L
<༐’
༐
༐

ʌ


༐


༐
༐
༐
>༐ .
QLų>༐% ༐
༐

@  ༐@༐.

 ༐




༐ Q
༐Q
༐
d9
༐༐E༐

༐
Q

༐“༐༐Q99‚
 ༐

༐ų༐༐Lų
< Õŕ Šʴ༐
΁
༐ ༐ 
༐ 
Q

༐ ༐ 
༐ 
༐ QL༐

༐ 

༐ 
‚

%
༐ %Q
༐ E
༐ 
E༐ 

༐ 9ųă<༐ Ø L༐

@ ༐ ༐ ༐ 


༐

༐J

༐%
.༐
༐%B
}9༐

9
༐༐ 

༐9

༐༐༐Øѵ༐9
༐
9%

༐ ༐ .9༐  9
༐ 
༐
 ༐ ༐ 
༐ 
༐ 
༐ ༐ 
༐ 9%

༐ .
L༐
%B
༐ ༐ 
༐ ‡ƪ7ºX}
Q
<༐ Œ༐ 


༐


>༐ 
൹
?“ ༐
ȇ
༐ 

༐
.

༐
Q

༐༐ŽXx༐. ༐
༐
༐༐
 

ų
༐༐X ༐ƛ

༐
’B
༐༐ Ŕ%L༐ƛ

༐XĻ τŠȪ༐ Ø

L༐ 

༐  @ Ȓ༐ 
༐
99
༐
༐
༐༐༐
༐9 ༐
༐

@ ༐

༐@༐
9%

d9
༐ Q
༐
༐ 
ų༐

 ༐>༐ ༐ 
༐ ų
༐༐ƪ77ºX>༐
༐  ༐ %B
[9༐ 

‚
9ų

༐ %Q
༐ ‡77ºX༐ ༐ %L༐ 9

༐ 9 ༐
T 9
<༐ ¯ ༐ 

༐
?
  >༐




༐ ༐ .

 ༐ ༐ 9༐ 



>༐ ༐ 
༐
T
༐
 ༐ ༐
༐ 

༐9

༐ ༐
༐ 
༐ ༐ 9
༐ Q
༐ %

༐  
Q
༐ ༐
 
༐ 
Q

༐ 

%
༐ %༐ Ś
B
༐
༐ < ŕŠǐ༐ Ø L༐ ß
༐ 


༐


%
༐༐
༐
༐༐
༐
T༐ !
༐ Á<༐ b a7©༐ ÇȔ
ų༐@?
༐ ༐༐
 E
༐ %
<༐
x༐ 
༐  Lų༐
T

༐ ༐ 
༐ 

9
༐
ªL>༐ Qă༐ .[LE ༐ 

༐
Q
༐ 
L༐ %

༐ 
Q

༐ %༐ .B

༐ ༐  L༐ Ś
༐ ØLL L༐ ‫ڀ‬%ʎ


ɂŠǬ༐༐
༐ų
@༐ ༐
}

༐
QE

>༐
9; ༐
༐ৼ
J༐

༐༐



>༐  
༐ ?
༐ 9༐ Lé
?
༐ ༐ 

;༐ 
9
j༐ Ø


༐


Ĭ༐ ;
E
༐ ༐ %༐ ༐ 
༐ ET9

༐ ༐  9>༐ 

;>༐ ༐

99ų༐ .  ༐ 
༐ ʄ9༐ ༐ 
༐ 
༐ ༐ a77༐ ༐ ‡77ºX Ɗ༐ 
༐ 
༐ Q%
༐ E༐
.
༐༐ų༐
༐L
༐ 

༐ ༐ .
༐ %9༐ ༐ 

༐ %
༐9ª
༐
%༐Q99༐

E༐
༐ E༐ƍJ༐༐
༐
༐༐ y༐ “J༐ B<༐ȭ༐༐ª[
7ϩ Av

7
ေေɊ
Q
ေ
Ƿေ AǷ
%ေ 2ေ W
(-Ïေ4-MေX
ưေʜĬေ'Mေ ๚٠ᛦ ·ေ
c4ԓ Ŧ 
^ǯ §bǯ 
¯“
ǯ Ú ǯ

 ǯ "
ǯ

 >༐ "
༐ .
༐ ༐ 
༐ ༐ 
!ś༐ 

༐ 9([ª༐ 
༐ ༐


Q
༐
༐9W༐ ༐
༐
Q
༐"༐%
༐
༐༐

༐༐
༐œ?
༐
༐༐ Ĭ  ǯ༐˜7༐
ˡĐЈ༐!༐
༐.? .༐༐!.
༐༐ª༐༐༐
"

!9
g༐ ]
༐ 
༐༐ª
༐9༐༐Φ༐(༐  (
W༐ 9 ༐ !༐
 (>༐ 9
༐ ༐ 9
༐ ༐ "༐ 
Q
༐ ༐ T;༐ Q!༐ (QQ(༐ 9 ༐
(" g༐ ‰!༐

?༐ 

༐ ༐
༐ 
;9;༐
Qǘ
༐
༐ %

༐ 
(༐ ༐ (TÔ
༐ "༐ 9ǯ@ 
!?Q
༐ 9༐ ༐ 
Q
ś༐
 g༐

ྔೆೢೆྔೆ໵౎ೆ྿ᛦ

ေ Ň^ေ Ǯေ ë Mေ €ေ Ɋgေ ř[ေ

ေ Šေ Ÿေ C
33x ေ d€
/x
ေ #
pƚ/ေ #
8/
ေ ေ
p
//ေ`[ေ

pေ#[#
/   o ေąԱ1ǯ ƓL 1ǯK 1 u ǯ/ ' —ͯ'‫נ‬ေ ɫ4
Çေ0  ေ
—ေ ”ေ ȣ  ေ4’
:3  ေdŢ [/x#
ေŢ3
3x
/ေေŠy
//   o ေ0b' eေë
ေ+
3
ေŢ3
// ေ
ÞǺę3 ေ 0ေ ëɏ
ေ 
/#3xေ ေ ‰
ေ
Ǻƚ3
/ေ /ေ Ƴ8
ေ :‘‰ ေ ‰
ေ ƚ3Ŕ
//ေေ+y
3
ေ3
//^ေ
ေ Ňေ Ǯေ ëy  ေ dW

xေ`
‰:

ေ

/‰#ေ 
ę3
`xx[ေ
ေɏƚ3
yေ
Ǻ
/ေ
ေp
//ေေ/ေ


yxpေ3
p
MoေąԱ  1 ǯǒ ǯJ 1ǯ '/ —2઄અ— ေ ɫ4
gေ 02' ေ
2 g ေ Ň g ေǮ^ေ ëေ
ေ +ေ Š  ေŐx# x ေ d¯
3
/ေ#
8/
ေxေ

xpေ#83v
/ေေp
//ေ`[ေ


ေ3

‰
 Moေ฾ ٓ ᛦ™ µϭ੢ & ǯ  0 ǯ/—I'b hေ ɫŇ3ေ0    ေ
 g ေŠ  ေa  ေC3
[ ေ dëɏ
ေ3ƚ3‘
/ေေŠ
// oေ—ေȨg ေИƚɏေ8`/ pေ+3 g   ေ
̠
:ေ ʜ3Ĭ ေ 0‫נ‬ေ
'ေ Š  ေÞေ”
/ ေ Š¯/#/[ေ
ေ3

‰
ေ3
3
/ေ ေp
// MoေW
/ေ®L3Lǯ (
 0 ǯ
 /  b'—ေ Ô02hƐ20 ေë
ေ [/#
ေ4#
‰[ ေř^ေ
Iေ Š Ç ေ‹zေ ”
/żေ dŠy
//   o ေ” ေay
[ေ §ᛦ 4/  ေ ‚#  „ ေX
:ေ·ƆĬ ေ 0'ေ
hgေ ȣ^ေÑေ+Mေdɤ [/x#/ေေ
ေpy
//[ေ/‰

„oေm83ေ
#83
/ေƚv
3
ေ
ေ
ေɏေ
Ň8
ေ +Ƴ8x8ေ ေ +yy
p
ေ  [/#x/‰/ေ
ေ ɏ
ေ 4
‰
ေ ʛv
3/‰[ေ ေ ‚ư
ƒေ
”8
ေ0ေ
ေ—% 0 †ေW ǯေ฾ ٔ ᛦ (
J
0ǯ / 2 ջ Yေ —ͯ2—Y ——2ͯ— —ï આઇ @ေɫm
 z   ေ
C
3  M ေŇ3   ေC
[ေ 02ေ
h
ေ {ေ 4

p ေŠ€
ေŠ
/ę
`3Ĭ
ň [/Ĭ
yx/#
ေ8ေ+
/#
ေŠ 38
p
 oေ
43p
3À¯
3
p ေɊ
3 ေ0Iေ ɫW
vx
:
ေxေą  ԱǯL ǯJ  0 ǯŇ8pေñ 0h ေ
0gေ {  ေᙿ‫ڈ‬ᛦ W
 ေd‫מ‬xŶ
/ေေ
ေŒ#xv
ေ

3
83
ေ##
 Moေą«Ա ™ ࡥ Ϯ ੢Ǔ ǯ
J 1 u ǯ'/ 2 ˜2'ေ ɫ€
# ေ 0' ေ
ေ WÇေ a  ေŠ  ေa[#ĬĂɌေ ൗေŠ  ေaေ C3
[ ေ dКĨ3
[ေxĂɌ3
#ေƚ
nj˶Ҡ3
ƚ/ေkေ /
ေ
/
Ĩx
Ĩ/#
ေp
//
/   o ေą«Աǯ)G

ǯG  0 ǯ'/—'b—'@ေ Ô0— ေ

  g ေ”  ေë g ေW

yMေ { g ေ  ေWĬ/`[ေ
ေ Ɋ Ç ေ 4Çေ +
3 ေ dë
ေxĂɌ3
#ÛxေေǺň3
[/ေ`[ေ
v3
8/ေ//ေ
ေ /ေ `

3xpေေ’
3ေ#/8k   o ေ +nL0ǯ   / 2hေ ɫC
3ေ
0 ေ ɫ48

ေ    ေ
ζ —  ေ ड़ ِ ᛦ ಶ ِ ᛦ a
33
kMေ dКO3
[ေ 

3
ေ   ေɏ
ေ /38#‰83
ေ  ę ေƳ8/ေ
ေ p
//żoေ
฿ ٖ ᛦ%%Gǯ (
1 u ǯ&  /'2ۣۤ'2ေ ɫ‹#ေ0 I ေ
 ေ Ɋ^ေĮ g ေa
33
ေ
ေŇ  ေ€^ေř3xp ေdഋĨ3
[ေɎ3
#xေ/‰8[ေęေɏ
ေ/Û38#8Ɔ
ေေ/
߽
/x#
ေp
//   Á ေąԱ 1ǯ L 1ǯ  0 ǯ&/ —'0ӕ—I'ေ ɫ4
ေ 0  ေ
2ေ Ɋ g ေŸ^ေ a
33
ေ
ေ ”^ေ Ɋ/#
 ေ Šm83
3ေ

y[//ေ ေǺÀ3
[ေ
‰
‫أ‬/ေ ęေ /
Ĩ/y#
ေ
p
//   Á ေą « Ա ‚ 1 ǯL 1ǯ 1 0 ǯ / —0b—'ေ Ô0 h ေ
Heေ a  ေ{ေ ‫
ݠ‬#
3x
/
 ေ dë
ေ
#ေ
33
pƚ
ေေpyƭ//   o ေą«Ա W ǯ ǯ= 1 0 ǯ /
h2ѕ hေ ɫȦ#ေ 0 — ေ
'gေ ‹ Ç ေř  ေŇ
3/  ေ ŠŸĂɌ
#ေ ęေ 3
‫׬‬π˶Ҡ3
ေ ေ p
//ေ /Ŷ38#‰83
  o ေ ą«Ա  %%G1ǯ (
 0 ǯ #/
02 β02@ေ ɫŇҦpÇေ0'   ေ
 )ǯ ggǯ

'ေ Čေ aေëÜေ dX([Äေ Ğ(ေ 
ေေÄy(ေpƭÄÄ Áေٍᇨ‫ي‬ᛦ 

 4Kµǯ Ƥ U;ǯ &  '  )
4ใ߀Ä ေ '/ ñ Gb2ေ FŇRpgေGIေ FǥޗIh—  g ေ
'` e ေČ  ေaေ ë- ေ dǥ³ေ   ေvĞʡ<Äေ   ေ³˳ေ ijŻ͠Ŷ5ေ `ʈ Ę[ေ py‫ݪ‬ÄÄ
Ä   Á ေ ًᇨ ‫ي‬ᛦ >•
 4iǯ
‘ U; ǯ"n4ǯ+ *# 4#
ǯƈǫȊေ b͇ϴ'ɽေ FŇÇေGhေ FǥޗhѠ̓2  ေ
Iေ Ţ g ေܽ Ç ေV [ ေ{ e ေǥ:Äေေ” g ေǀ g ေ4Ŷƭ:ĻĜ  ေdX
:ေʡӅΣ
ÄေေpÄÄ5Ä ʸေ‘ UM4Ēǯ
$%/ GઈઉG ေ FȦ(ေ—I  ေ G' Çေ
hÇေ ࣴûَᛦ ř g ေV[-ïေ{ g ေǥ:Äေေ”gေǀ  ေ4Ŷƭ:ĻĜ  ေdȀ
Ä
ƭ( ေေ5:ေpÄÄ
Äေ`Ä
ေ
ေ-
ေǺĞ
Äေေ ʡ5yRĞR ေvR ေ³[`
R ေေŶ RĘpÄ-
 Ç Á ေë 5ေ Ţ‫ف‬
(

pÄေ ေ -
ေ ǁRĻ- ေ ‚Ŷ
‫أ‬Ą Ğေ +p
ÄÄေ ေ Š³ƭÄÄ ေ ŢƭʈÄ ေ G'Çေ ŇÄ Ǵေ
d4
Ğ(R(-pေ<
Ĝ
Äေေ Ä
ေ vʡ
ေp³ÄÄ
Ä ʸေ ฿ٗᛦ
;Ɣƴ4=Ƶǫ¹NJ2ǯ =K 2 ‰ ǯ
ŽǫƨȊ — Ib— Gေ FŇေ G   ေ
V [  o a Ç ေŇ Ç ေa
[y ƺေ d+³RĻ5ေ Š³ÄÄ
Ä ƺ Á ေë
ေ 4(
[ေ ေ Š³ÄÄေ ë5( pą ေ 4
Ă‫؀‬5 ေ
Ůp ေ G gေ
—gေ ǁေ ġ Ž ေë
[Yေ dî`űေေŠȿÄေ‫܈‬ౝR(-RĻ5   т ေ Ӿp5ေ ŢR`³Ä ʈĘpေ+  Ç   ေX
:ေ
Ǧ<ű Yေ G ေ
— gေ Ňေ ವ‫ى‬ᛦ ŏƭ³5ေေ”ေǀေ 4Ŷ؏ؔ:  ေ d4
̨ေŠ‫ݲ‬ÄÄ5Ä ʸေ ฿ ٖ ᛦ =Ƌǯ );

ǯ K*=; 2  ǯ' /
IIbGေ FG2h  Ç ေ
——Çေ Š Ç ေŠ
y ǔɍေေCgေë Ä ေ dë
ေ [ÄĞ(ေ5<-
Äေ  ေpÄÄ
Äေ ຟေ 5ေ
‫ؽ‬


(Ȃေေ(ÄÁေFʈေŠ
ƭ  ေ-ԱKµ 2ǯ µ(
2  ǯ#  /—'ઊઋ—IhေF”R
ေG—'   ေ
— ေ Š  ေ4gေǁR(
Yေ dŇyӅΣÄʈÄေ  ေ5(5-ေ 5ƭÄR55ŶÄေ  ေĜ
ေ vÄ(ÄĞĥ[ေ  ေpyÄÄ
Ä Áေ
฿ ٖ ᛦx LJǯ4 ǯ=K  d ǯ%/ GÀα ေ FӸR
ေʽG— zေ
—ž ေ ǥေ aÇေ സ‫ڀ‬Rp‫ݲ‬ÄYေ dë
ေ ‫ޚ‬ยÄ
ေ ေpyÄÄေေ p Àv(RRေƭ‫׷‬ÛRÄ   Á ੥ Ç ေic2 Ơ
ð ǜ V s ǯ̉̉ Ȋေ
hbhIေ FC[ေ Gɽ ေ 45ေ ƭ-ေ 5ƌÄေ :ƌ<5ေ 
ေ ေ Ą ʈÄေ Û`5ေ `ąေ Ë
ေ

Ä5ေR ေေŻR
ေ-ေű³[ေÄRĞ
ေ`[ေŶ
ေV-Ä ေë ÄĨîRÄÈ
-ေ +Ç ေ ČÛေ
—ေ ǀgေ Vေ4 ƭ ေdŠÄÄေǀp55pေ{ƭ`˳   o ေ—ေ
 Ç   ေC(Šƭ:ေనေVűေ+ ေ
‚(  Y ေX
:ေǦ<ű ေGhÇေ
—'zေ ǁေ ”gေ X-ljေ dƼ

-ေေpÄ5Äေ Rp ေÄ³Ä   Á ေ฿ ٖ ᛦ x %%;ǯ (
2  ǯ% / 2b Gေ
F”ေ GI gေ
—I ေ ǁေ ”gေ X- ေ d{
Rေʈǔ෎RÄʈေ Rp ေp³ÄÄ ʸေ ฿ٖᛦ ఆ ፓ ٗᛦ ß4 2ǯ=J 2 d ǯŎƈȊေ GbG'ေ
FCေ G   ေ
—hေ a ^ ေŇ  ေa
[³ေ d+
(ေ ȿ
(ʡÄေ   ေÄ
ေ 5( ƭ(ေ <
-
Äေ   ေp³ƭÄÄ   Á ေ ï

 4Ksǯ ̉Ȋေ ‫ 'ۀڿ‬ေ Fҹգ2h gေ
—G ေ ǥÇေ Cgေ V
 ေ d€ĞɎRÄေ ေ ေ Ĝ Rp ေÄʈÄ   Á ေ+`<p
ေಜࢍĞv5Ä-ąေ
ÄÄ ေ
+ƭ`Ğp
Yေ G ေ
Çေ Ţgေ ܽgေ 4ʡ ေƭေ X  ေa g ေë[³ ေ d ŏȓɍRÄҕေေ
yĞRေ <Rp ေÄ5v
‫׷‬ေpƭȿÄ5Ä ࢌေ
฿ ٖ ᛦ ¨ǯ 4 ǯ =K 2  ǯ̉ŎDZ  Gbñ ɽ'ေ FCƭ[ေ Gɽ ေ
ş gေ ëေ ǁ  ေဣ5:ű<űေ ေ ǁေ ġေ ë5[ ေ d4Ŷ R[ေ ေ
Ɏ5(-ေ ေ(ƭ-Ğေ((
Ą ƭʡĞေƭေ
ÄĀ
ေေ
Rေ5
`³-ąေေ˳³ĞÀÄĞ(ƭေpƭÄÄ
Ä   Á ေ฿ ٘ ᛦ   2 ǯ4 ¸ǯ =K   ǯŎ̉ ýေ
—I—b—Ihေ F4
 ေ G2G  g ေ
— g ေřေǥཾཿ[³
p  ေ dŇÛ-
ʡေေ
-
(-ေÄÄp
ေ  ေ
yĞRေ ɨRp ေေ߉[Äʡ໻ေ--(
ေ
ေ p ေ

-R
Ä Y Á ေ 4=K ǯ =(ǯ= ǯţǭ=*#= ǯ )/ ŽƈŎDZ I'bνθhေ FG I ေ
gေ a Ç ေ{ေ ›

Ä ေ d{
R   ʸ ေǀyÄ5vʈ5ေſűေ+[  ေ ‚(    ေX
:ေʜĻű ေGۡ—gေ
2gေ 4Rေ€ԈЋ  ေ d4(ʈ
ʡȓ‫(ޗ‬ေǁRϒ˵Äေေġ(RRေë
( ĞƳR5 Áေ” ေa
[ေ ¨ᛦ
4ÄYေ Ŀ( g   ေX
:ေ ʜű  ေ G2Ggေ
ေ ”gေ €ေ ǁÄ- ေ dŠÄေ
<
ƭ`‫ޜ‬-[ေေ5ʡyÄ   Á ေ c;c%
ǯ K2ǯ >o , ‰ ǯƈ DZ 'bI ƒ ' ଢଣ2ေ
FG2  ѳ ေ#/ @2ɸh—ေ FG2— ေ
'Çေ X g ေ aÇေ ë[Ļေ ေ a³³ေǥÄ ေ d ˃ĞФRÄေ ေ
Rေ ေ [<p
ေ - <Rp ေ
[
uေ `ƭေ ( 5([ေ 
ÄʈÄ-ေ pߙྀÄÄ   ʸ ေ ฿ ٙ ᛦ  2ǯ ®(
 0 ǯ ƈȊေ ' ӟ —b'Gေ F‹(-ေ
G hေeေ
cfԓ P 
Ŗǯ W :ǯ 
/
ǯ . ǯ

 ǯ ŕų
ǯ

‫ۼ‬ʪI Ç ေ aေ ΃ေ Ñ[Yေ d²8¦
ေ -8*
ေ*ေ


ေ p

 ŭ ေ
5
`**[ေေ v
ĞRေ p
Ä
ေ -ေ
5*8 Y Ɣ ေ฿ٗᛦ ǯǖǯ  0 ǯ/ hhIb 0ေ ÔȦ#ေ0 —£ ေ
؈ gေ agေ ౱‫ك‬ᛦ Ѧ[Yေ d М ़*ĂɌR*ေေ [ߐp
ေ
ေ 5*8ေĜ 8p ေ *y*#
ေp]ÄÄေ ]ေ ĥ
ေ
p

 Yʸေ฽ ‫ ل‬ᛦW   ǯ¹Njǯ”   ǯ / —'Ϥߺb'Iေ Ô0——£ gေ
0 ေ + g ေ +  ေ ġ]ေ ġ *Ä èေ d΃*ĂɌRÄ*ေ ေ
*Rေ Û Rp ေ 
v
]ေ :Ğ
³[ေ *Ɏ5*pေ
p
Ä
 Y Ɣ ေ (
ǯ o   ǯ)/ Iေ ÔӪϐ
[ေ0—2£ gေ
2 gေ řgေ 4ေ |‫٭‬4
*ေ
ေ |ေ Wေ {p
 Yေ d ŏ*ĂɌ8*ေ ေp

ေ Ë 8p ေRÄ
ေ Ƴ8
-Ā Y Ɣ ေ ฿ ٗ ᛦ
(
ǯ ¹nj V  ǯ / —0ɸ—'ေ ÔÞ#ေ 0 —£ gေ
2 ေ lzေA  ေa*y³*
ေ
ေ”ေV Ç ေ²
p8njಹYေd4R`**[ေေ
*Rေ
ေ [p
ေ*ေ

’
ေ
**#
ေp
ေ
ေ-
ေ 
y
-*ေေ Ë *ေËေ
ေ


`**-[ Y Ɣ ေ฿ٚᛦ    ǯǯƃƶŁ   ǯ
   /' Ϯ 0ေ Ô0—2£ Çေ
2—ေ agေ Agေ Wp
YေWေ 4ေ VR¦*ĜĶေ
ေ ౳ ‫ و‬ᛦA
-Yေ d ŏ*Ăɍ8*ေ#
ො#*
-Yေ8`*³*-[ေ
ေ



`**[ေေ 5*8ေ*ေp
 Y Ɣ ေ฿ٚᛦ ¦ %%Gǯ (
  ǯ/ h'hbIေ ԔR[ေ 02£ Çေ
2ɣऻ  ေ Cေ ‫܍‬Çေ W
[yÄYေ d€*ĂɌR*ေေ
pေ *ေ
ေ ĥ
Ä*8Ĩ*
Ĩ**#
ေp
Ä Y Ɣ ေąԱgေ Ĥơ¸ǯ
& ǯ  0 ǯ   / 0b 0hေ ÔXvေ ľI£ ေ
2
gေ › g ေV  ေC#A55Yေ”gYေ d4ĥ 5Ĩ5
#
ေ *ȓɍ8*ေ *ေp
Ä Y Ɣ ေ ฿ٚᛦ ç  ǯ (
V  ǯ%/
—hɸ ——0ေ Ô²
` gေ 0h£ Çေ
22gေ 4 z ေ4# RQ:YေdȦေ
ေ*‫؂‬R5#5ေေ
``5ေ8
#
ေ
[5ေေ
ေ-
kp ေေȓ*
ေ
**#
ေ ‫؂‬ร`5 ʸေ Ô*ေŠ

£ Yေ (
$kǯ «Ա ƨ+M$ǯ / H— Àɂ ေ Ԕ]ေ 0 —£ ေ
2ေ j$æ ‹:
Yေ dX# ေ`*--

ÄÄေ
ေ-
ေÄ-
pĥ ေေ
ཱྀ
ÄY Ɣ ေ î 
ǯ
Ʒǯ
3ǔ
ǯ
$%_ Ǟ$#4
ǯ$ǯ+  d ǯ̈́șȊ İ'˜—ေ Ô΃
# ေ 02£ gေ
2'ေ Wgေ Á ! ͕›# Ûေ]ေW Ç ‫ט‬хေC*Q

yYေų4Û
pË ေ
ေ
*pR
ေေp
ေ
‰ေv
[ေ:ေ
Ë


8Ν
 Y Ɣ ေ฿ٚᛦ W %%Gǯ (
  ǯ'&/ '‫ܢ‬ї'2ေÔC
[ေ0I £  ေ
2IÇေ AေAေl*‫ *ކ‬Yေd|
ေ


ေေ¦ྖ8¦
ေ
ေ‫؂‬:ေ*ေ* Y Ɣ ေ î 
ǯ (ǯǯ
ƥǯ)/ ̉̉Ž ý ' b0hေ ÔȦ#‰ gေ 0—£ ေ
2h Çေ Ĵ ϫ ੢S î ੢ř[# ေ
ေ²ေ ġŽေƲ5[Yေ dǀ‫׿‬ę5#-ေေ-
Äေ
ေ‰


8¦
ေR¦*pေ*pေ
ေÄĥ
p- ေေp
ÄÄ Y Ɣ ေ฿ٗᛦ ǯ& ǯ   ǯƨǫý IbH—ေ ÔAေ 0I£ ေ
ž0ေ ǀ  ေÞ:
Yေ d|
ေ*#

Ä
ေÄ-
pĥ ေေ *ေ Œ`
Yေ Û
ေ ”ĂɌ‫
ݿ‬ေ
ȓɍ
#ĥ Y ေ
ေ 

ĥ
ေ
҉ϓ
5
ေ*ေ
ေ*p -ေေŠ*͜Ë ՘ေĥ
[ေေ
#8¦
Ɣ ေÔ*ေl5
£ Y ေ-Ա X
  ǯ
% / 0ଚռ— ေ ÔÞ#ĥ gေ ‫ی‬0 £ Ç ေ
ေ lgေ²  ေ4-#Q

Yေ² g ေġ g ေ|
[ေ
ေ+ g ေa Ç ေ·R*pYေ d+
p
ေ*ေ
*
ေ-
p- ေေ
p
ေ#
R
ေ`[ေ:
Ą
ေ*
*ေ-

ĥ
YƔေ฿ٚᛦ §ǯ& ǯ  0 ǯŎƨý  'b—ေ
ÔA҉ϓ g 0 £  ေ
 ေ ġ  ေ›  ေC  [Yေ² ^ ေġ  ေ|
[ေ]ေl g ေ²  ေ4 -#Q

Yေd‚ʟR
#5ေေ:

ေ*
*ေ
-


ĥေေ
*
ေĥ
p- ေေp
 Řေǀ‫׿‬5#ေေ


țR¦
Y ʸ ေ฿ٛᛦ §ǯ& ǯ
   ǯŎșý 0b 0hေ ÔXvgေ 0'£ Çေ
—gေ +ေV]Yေ ”&ǯk3Gǯ)
ǯ °$

ǯ)Ƅ++$3ēǯ gေ22ေ ɫş෩02£ gေ

 Çေ { g ေ› g ေW*#
Yေ d48 8ေ*ေp
ေ ေp
Äေ:˳*p Y Ɣ ေ )G

ǯ #   ǯ/ 0Yေ ۡ'ေ

Ԕ85ေ 02£ ေ
2 gေ ”Çေ ฮϐ*Yေ d4
ေ
ĂɌ
#Ĝေ ေ p
5ေ ေ  5**#ေ Ǻ*
À#
ĥ
ေ #
- 5  Y Ɣ ေ
W
҉ϓေ   ေ ĥ ေ AR
ေ +
5#
ေ  [Ä*#]ေ ȣy
#-*#ေ    ေ —b—Iေ ÔC]# ေ
0£ gေ
gေ W g ေa Ç ေ΃8py
ေ
ေ ”ေ Ȧေ ‚
Yေ d|
ေ
#*ေေ:
-
ေ]ေ ေR Rေ*Ǻ*
ေေ
p
ေ R
#
ÄYƔေ฿ٚᛦ ǯ)G

ǯ G   ǯ / ࢎࣥ‫ ڹ‬H2Y —h0b ɽ0ေ ԋ#ေ 020 £  ေ

' gေ Aေa*
Yေdl³
ေ
ʡ* Y Ɣ ေ฿ ٚ ᛦ ǯ& ǯ   ǯ/ 2bhေ Ô²
`gေ0I £ Ç ေ
I gေ |ҳ*p˙48*ေ› ବ „ ေd௸Ǻ
*

ေ
v*
#5ေေ v*#Rေ`

v*ေ  ေp̡Ğေ`R
*
ေ
Ğေ

 Y Ɣ ေ (
ǯ o V  ǯ$/  ˜ž'ေ ÔAÇေYေ ñľ2I£ gေ
I
g ေ Wေ l߭Yေ d+
#R
Ą*ေ ေ ĥ


8
ေ **`8ĥ*Äေ *ေ p
ေ 
-
ေ R
È
p*pေ

ေ-

-
 Y Ɣ ေ฿ ٙ ᛦ W   ǯ& ǯƅƸý ǯ/  / —࿶ਘ—0ေ ԔR
ေ0h£ ေ
h ေ l ^ ေ€ေ W
-Yေ d| 5ေ #-ေ ေ
**#ေ v
v
ေ
v

ေ Ƴ8
*-[ေ`[ေĥ

ေ
 Qေ
*pY Ɣ ေ
Gƌǯ
j   ǯ% / 2I˜2Iေ ÔXvေ0' £  ေ
 <¹

Ա )

4Ա ˆ
%ေ{ေ W    ေdV#ေS#S
ေေေ
টćေ <EԱd1  j Աbe BԱ0ƒ08 ŌԱ08 Ա

Ԕ8Ƅေ œ4D… Ա
­D Ա ž͕ {  ေAĢေĢေj ' æĸ  ေayy %ေ dë
ေ 
ေေ % ć ေ฿ٖᛦ ™)dqE>dԱd f ̉ Ա
'/ 4Iș­0Ͷ !Ա 80ū8IDԱ Ú`4 D… fԱ
­œ Ա ž͕ {ေ AĢ%ေ dë
ေ
ေေေေေS Ƅ#ေS7
„oေ ฿ٖᛦ ™)dqE>dԱdԱ ͇ Ա
ŐetB 04̷I`Ա ԔRƄေ `400… Ա
­ Ա mေ aေ 
‰%ေ dAေ
Ĩ
Ǻ-ေေ
ေAĢေ Ģေaေ:ေ ေ
Ŕ
 % o ေ฿ ٖ ᛦ
&ĀԱ<EԱ &dE f É Աbt BԱ 8Iƒ 4­Ա ԋ#ေ œ4 … CԱ
­0 Ա mေ ë:Ƅ%ေ฿ٖᛦ
&Ա<§Ա  &dE  ! Աe B Ա­œƒIJԱڜ4`I… Ա
­JԱ { g ေ W g ေ ˆȂ%ေ d4
ေ 
y

ေ ေ „ေ ေ S

ေ vvေ v#-Ƅေ ]ေ ေ
v`y
ေ: ေ
% o ေ฿ٜᛦ
±Ա )±Ա
&  ! Ա'/JūJԱ \ӯ
7 gေ œ40­ũ ÊԱ
­Ա Aေ Ǯ g ေë %ေ dW
Ǻ‰ေေ-

ေ ေ 
ေ „oေ ฿ٖᛦ   )&Ա EԱÁ) Ա

d1)1jԱb– BԱ`44̸ `œԱ \m
7 gေ œ4J… fԱ
­­ Ա lÇေ ĸ g ေW
Ģ‰ေĢေ ”ေ ಶّᛦ 4‰: %ေ dë 5ေ - Ȃ
-#ေĢ
v
S
ေေSŒ
Ģေ

ေ#
Ģ8
 qoေ฿ ٖ ᛦ
&Ա<EԱ &dE  ! ԱbŐ BԱ0 ½04Ա \m
7 gေ `4J8… fԱ
­I Ա Aေ ǀ  ေĸy
ေ Ģေ ” g ေǀ  ေ4‰:-  ေ d‹ေ 
ေ v#‰Ƅေ ေ 
ေ 
ေ ေ -
ေ Ի

ေ
% ć ေ฿ ٖ ᛦ
&fԱ <EԱ &dE  ! Ա'/ J`JƒJ IԱ \ĸ
#ေ œ4J… Ա
­8Ա 4ေĸvĢေſ
:5ƒေ dWȂ8‰ေေ
ေ 
#
-ေ
ǺS

ေေ-
ေSS

ေŔ
S

Ģေ8Sေ ေ7Ƅေ‰
ေ#-ေေေ
ĢေေĢĵ

ေ
S58¦
ေĢေ
#
Ģေ RĢ
ေ Ģĵ

ေ ##8-#
 % ȝ ေ F>E Ա )dԱ *EԱ
& Ա Ćd1d!Ա
J0­ƒJ04Ա Ú`8œJ… ɆԱ>n>1  j ԱX8
¦Rေ
ေSS
ေ7Ƅေ-
ေ
ေ8 %ေ Ú`8œJ½œ8œ­… Ա
­4 Ա ”gေ {  ေĢ
%ေdlĨ˿C
-ေ4
%oေ4#
Ƅေေlေë
# Ƅ%ေ4
Ǟ
Ģঠေ
ǀĢ%ေ œ4J4Ա
IDԱ ๻ ‫ ف‬ᛦ {ေ AĢေĢေ –ေ ĸေ a%ေ dë
ေ
-ေ7
-:

ေ7

#
ေ ]Ģေ

ေေv8ေƄS
ေေ „ o ေ฿ ٖ ᛦ Ŝ>d/dԱ
&1Ա
&> j Ա'/ ­D4ƒ­ Ա Ú`4 ȫ 4… fԱ
I ` f Ա Agေ Aေ ɤĢေ Ģေ ”ေ Ģ
ေ ¯
 %ေ d4

ေ ေ ေ Ģေ ‰
ေ 
8¦

 „ ć ေ
F>E>sԱ &fԱ ‹  ! Ա'/ II½ 8JԱ ڜ4JI½œ4J8ũ f Ա
I Ա +  ေĸ g ေ4S
#
ေ]Ģေ4  ေ”
%ေdĸ
pေĢေ#8#-ေေS#SǶေေ8
ေေ
ေ#‰
 % ȝ ေ฿ ٖ ᛦ
±Ա )±Ա
&  ! Աe– BԱœ ȚœIԱ \”8Ƅေ œ40œ ũ f Ա
I0Ա ”  ေm‰
Ƅ%ေdAေSǧ
#ေ‰ƄS
ေေĨv
:
 % ć ေ฿ ٖ ᛦ
&fԱ<EԱ  &dE  ! ԱՑ B ` JȚ
œ IԱ Ú`4J… fԱ
IJ fԱ W  ေaေ l]ေ ]Ģေ ˆ  ေ{ေ AĢ%ေ dAေ 
 Ģေ ေ -
ေ S
#
ေ
8

ေ ေ
S‰#ေ S ‫ڤ‬Ģȓɍ

#
% ေ
S
#Ƅေ ေ ‰

Ģေ ]„oေ ฿ٝᛦ ™)dqE>dԱdf!ԱŐetBԱ
JIƒDJԱ Ú`400 …  Ա
I  f Ա a g ေ ë  ေW
Ģƒေ d‹S#ေ 
R¦

Ëေ ေ
Ģ8ေ 

/ေ ေေ 787 % ć ေ EEԱ
nԱ &)1!Ա% / ­ ƒ­Ա \ӯ
7 ေ `4D… ĠԱ ฿ ٖ ᛦ
ssEԱF*  ! ԱŐe BԱ Dū IԱ ÔCေ œ4D… Ա
I­ĀԱ Aေ Aေ ðĢေ Ģေ ”ေ Ģ5ေ ¯Ȃ%ေ d45
ေ ေ ေ Ģေ - Ȃေ ƌRȂ
- % ȝ ေ
F>E>sԱ  &Ա ‹  É Ա'  / II½ 8JԱ ڜ4JIȟJ4œ8… Ա
II Ա ¯  ေ+ g ေĸ
C%ေ dAေ:ɂ#-ေS7
ေS#S
%ćေ
*E‹d>Ա  &dE/>ÉԱ%/ ­JԱ
ÔAS ေ `4… fԱ
I8Ա mေ aေ 
%ေ dë
ေ 8
ေ ေS#S
ေ ေ 
ေ ေĢ8-Ƅ% ć ေ฿ ٜ ᛦ
±Ա )± fԱ

&  ! Ա/ 4ū­ 0Ա \4
Sေ `40D… fԱ
I4fԱ a  ေ ëေ W
Ģ%ေ dAေ S#ေ 
 Ģေ ေ 
Rေ -
ေ 
ေ ေ ေ 787 % ȝ ေ฿ ٖ ᛦ

ssEԱF*  É ԱŐe B Dș IԱ \Cgေ `4D… Ա
8DԱ Agေ Agေ Ģ%ေ dë
ေ
‰ေေ ‰
ေ 
ေ
ǺSေ#
͜#
ေေေေ
7Ƅေ 
ေ 
8ȂȂ-ေ ေ ‰

ေ SĢR#ȂĢေ 7Ƅေ 8ေ ‰ေ ေ ĢĢေ p Ü o ေ
<EƆԱŕd1   ԱbbB œ44Ա ÔASgေ `4 … Ա
8 œ  Ա ܷ l gေ +Q
ေ ]Ģေ ˆ Ç ေ{  ေlေ m %ေ dAေë
-‘
ေ ေ -
-‘#Ƅ% ȝ ေ+7ĢȂေ
Ñv
Ƅေ
 %ေ `40œ ÊԱ
8 Ա +ÇေCƄေĢေC  ေlȂ5Ĭ%ေdë
ေS-#ေƄ-
ေေS#SȂေေ85ĢေေS -Ŕ

#-Ƅ „ o ေ฿ ٖ ᛦ
&>fԱd) f ̊ Ա/II½8`Ա \Cေ`4J8… fԱ
Aǯ ʯ԰ö
ë

Ա
Ա + 5
Ա ™Ա ë Ա 
Ա

hé ေ Aေ aေ {8ေ

ĢေŸeေ Ÿeေ Ɋ8p
%ေ Ȉl
Ĩਙ

ေ

%ćေF*>&_Ա Ɇýေéh2Ӑ2ေ \€
( eေ
H0é2Ēေေ
hɽေ {ေ x‰Q[%ေ ȈA
[xေေ

ေ

ေ xေ 


Ģေ([xĢ
ေ
Ģေ 
ေ
ĂɌ
(ေေ
;x(Rေ ʟ:ေĢRxpေ


xp% ć ေF*Z&!Ա Ɇýေ2'β2H H ေ FH0é2 ေ
h2
ေ   ေC  ေ4RYေȈ4
ေ

RĻ

ေxေ(x(8
ေ([xĢ
% ć ေเٗᛦ
±Ա9)Ӆ fԱ
& f _ Ա
ƨŽý HéȋH0ေ \C
eေ H0h Ē  ေ
hေ m  ေ a e ေ‫
ן‬xYေ Ȉ4

ေ xေ p
Ĩ‰Ĩ

ေ 

ƛŭေ ‚ e ေ|
ေ ([³xĢx(
ေ 

 % ć ေ เ ٜ ᛦ

±ˆԱ9)±ˆԱ
& ˆ ! ԱŎ Ŏ ý ေ——2b——0ေ \”R[ေH0Ē ေ
hƭ ေ |eေ ŏ e ေ ǥx
[%ေ Ȉ 

‰x(ေ Ģ

x
‰xေ ေ 
xĢ8
ေ 
ေ xေ ‰
ေ Ģ
ေ ေ
ȂȂ(‰ေ 8`
ေ
;

 % ć ေ F)&fԱ
&fԱ Īs fԱ
)9Ա
dE  ! Ա ŽɆýေ X^  % ေ H'HӏHIေ
FH0IĒေေ
h` ေ ²eေ Ÿp
 %ေ Ȉ‹ေ 
ေ [

xေ ေ 
x

ေ (8;Ȃćေ \xေ l

Ē „ေ <E9&fԱ
ģ9) f É ԱŎŽ ý HééʺHéIေ \Aေ H0h ေ
h' Çေ A  ေ+eေ Ɋ
(Ƴ8

 %ေȈX
ေေĢR(xေ}ေ

ေ

(x(x[ćေ \xေm
( Ē %ေ ±sȜԱ
9dŸ f ! ԱŎ̈́ýေ0ေ FHh2 ေ
hI ေ aeေ | %ေ ȈŸ
([‰x(ေ (ĢR(xေ xေ xĢ ŭေ Œေ
Ǻ
ƛȂ%ေ ေ p
 Y ć ေ
F)&Ա* fԱ
& f _ Ա̉ Ŏ ý ေ2'h˜2I—ေ \”8
ေHhI   ေ
hh ေ ”eေ l
8 %ေ ȈŸ
([xေ 


ေ x  ေĨp
ေ 
ေ  ေ
(xෟ
ေ 8`
 % ć ေเ ٖ ᛦ

±fԱ9)±Ա
&  ! Ա̉ș ý —IIb—hHေ \‹( e ေH02'Ē ေ
h0Çေ ”ေ 
[Üေ Ȉȣ
([xေ


ေxေ p
ćေ \xေ m
( Ē fေ
ddfԱüŸ>ɵEfԱ &±ԝ
s/dfԱ‚)d&fԱ&fԱ ™f
f_ԱŎ ý HI˜HH2ေ \Aeေ H02hĒ ေ
0ေ leေ Ÿေ WxĢ
%ေ Ÿ  ေ +ေ C
`Ȃ
ေ
Ģေ a e ေ Aေ a
[ %ေ Ȉ‹ǺxĢ
‰xေ
Ģေ 
Ģ8(‰xေ ေ
pƛ

ေ`[ေ

ေေ

(ƛ[x  % ć ေเ ٞ ᛦ
±ĀԱ9)±ĀԱ
& f ! ԱŎǫ ýေéH2béñ0ေ\‹(ေH02IĒ ေ
0H ေ ๗ᛦ 4
ĢĢ
%ေ Ÿေ ”ေ |x
ေ
Ģေ aeေ Ÿeေ 4eေ |8
%ေ Ȉ|
ေ

(x(
ေ(ĢR(x;x[ေေ
ƛxRေ 

xx(

Ĩxx(
ေ pƛ

%ćေ เٖᛦ
&fԱ <EԱ 9&dEf!Ա Žƈýေ 2ࣵ˜2IIࣶေ
\H0é—Ēေeေ
0—ေ lေ l
ƛ Ɏေ
ĢေC  ေ| 
 %ေ Ȉ|
ေ  [x(
ေ
xȂေęေp

ေ
ေ
ေ ę8(xေ
ęေ
xေ(xxćေ \xေl

Ē Yေ CԱ 9&ԱF*Zq  ! Աƈýေ22ǡ2ေ \‹( e ေH0— ေ
0éေ ”  ေCေ 4
;
ƛ ေ Ȉp
ေ xေ 
ေ |
[ေ ေ 
ေ  [x(
ေ 
x
ေ ေ l
%ćေ
Ÿ
;x
ေR`x xpေ+%ေ‚(  % ေй
:ေ·Q%ေH02hေ
02ေ ๘ᛦ ǥ
( ေ
Ģေ²ေ aေ{xx( 
%ေ Ȉ‹ေ
ေ

‰xေ`
:

ေ
ƛ
(x(
ƛေ(Ģ8(¿x;x[ေ

Ģေ


RĻ
ćေ \xေl

 Ē Y ေԱE9q)&9±>9!Ա Žƨýေ2Hਚ2'ေ \”
eေH0hэ ေ
0ေ   ေřeေ ›x`[%ေȈŸ
(x(
ေ(Ģ8(țxေxေ p ƛ
 % ć ေP> fԱ เٝᛦ
ssEfԱF*  ! ԱŽ ý ေH0éӏ——ေ
\A8peေ H0Ē ေ
0'ေ ”eေ   ͕a x


Ģ%ေ Ȉř
(RĻ
ေ ေ €x
ƛ
(x(ေ |
[ေ
Ģေ ‫י‬8
x% ć ေX
:ေ ʜQɬေ
C(l
:À{xƛေɊQေ+e%ေ‚(  % ေH0—IÇေ
ࠨ౻I  ေ ๙ᛦ ”eေCRĻ [ေ
Ģေ 4ေ ‹eေCp
„ေ Ȉ|
ေ Ģx

(x(ေ 
Ą x
ေ ေ xR
¿xpေ 
È
‰
x
ƛ  % ć ေေ H % ေģ9EEԱ
*9'Ա 9&fԱ เ ٖ ʵ ᛦ Žƈ ý 20ébH—ေ FH0éIĒ Òေ >n>ĕ f ! Ա eေ —%ေ Žƞýေ'2Ɂ
''0ေ
0hေ ”ေ |ေ řx
‰ေ
Ģေlਗ਼ေaeေ C
[%ေ Ȉ|
ေ Ÿ
(x(
ေ 
ཁÛx
ေ ေ ͋
 % ć ေ XȂ:ေ
·Q%ေ ” ေax
[ေ §ᛦ 4Yေ ‚(  % ေH0éé ေ
00 ေ Ÿေ ষࣳ ‫ؽ‬ᛦ lR[
%ေ Ȉ|
ေ

(x(
ေ`

;xေေp
ေ
ေ ေ


8
% ȝ ေ เٖᛦ
' f Ա
ǜ)±fԱ
& ˆ ! ԱŽ ƈ ý ေ'Iё'Hhေ\€
(ေH0‫܇‬ϴé  ေ
Hေ a  ေA e ေa
[ %ေ Ȉ| ȂေĢxȂ
(x(ေ
‰x
ေေp
ေ
Ģေ
xေ‰8(RĻ
ƛေx-


‫ؿ‬
x % ć ေเ ٖ ᛦ
&fԱ<EԱ9&ſdEf!ԱŎŎ ýေ——b—éhေ FH020 ေ
Hȱေ ८ࣵ١ᛦ แ ١ ᛦ༝ ١ ᛦ 48‰Yေ Ȉ€x
Ȃඳx(ေȂ¿xȂေęေ;
xRေpʍ

ေ
ေ
ေ R(xေęေ Ó࿝
༬8Ók(߰ေ
kœေ ԳÓŵÓ
ĥ ྚÓٔေ ‫ٷ‬xkေ ఊÓ
Ē %ေ ‫ؿ‬භᛦᐵᅋጵ ‫ـ‬ᛦ Eђҥѓ²ҤŇшf ! Ա ̉Ɇýေ Iñ ÚbIေ ź‫ץ‬ħ‫ ب‬ေ
̙ΉͿͷ H э ေ
֖૖ ေ {^ေ ġ
x  ေ Ȉ4x
ေ
 Ģေ ęေ ’
ေ Ģ

x
xေęေ
ေ :

ေ Ăxေ
Ģ ƌ•xpေĥ  ေ
p
ßေ \xေl

Ē %ေ C ԱF*>q f Ա9±>9 f ! ԱŽșŎýေéIh˜éhေ\H022Ē ေ
 \Ć


Ա fǼԓ

)&5!o ǁ  ေǁR
%ေ
sO ,EÖԱB'
o FCJo FHJoH / &  oHLlo
)J5F  o
)&7 !o Wေ Č  ေ l55ေ
ေ ĉ  ͕ Ȍ͕ Ɋp
- Yေ Šǀ
(<*(
[ေ (R(ĥ *pေ p͠óƌ%Ɣေ ฿ ٖ ᛦ œ ʇѯѰԱ
 OʇҦҧ Ա
å,  Ö Ա8*
o ^J )&&o ÔC
<ေ + Jm ^  !o
)&n !o ›ေ  [ ͕ Ɋp
-Զ%ေ Š48ǰ
(
ေ (R(*vÐË[ေ ǰေ 
]ေ *³*(
ĥ5ေ p]ေ
ǰ
ေ ƪ[p
ေ

]
 % Ɣ ေ฿ ٖ ᛦ
'Ա Ӱ'Ա
åÇ - Ö Ա8:
);5Fo Ԕ
ေ )J@)   o
…2
ေ {  ေ{  ေǁ8Q %ေ Š‹ေ
ေ R<ǰ](
ေ (R(¿Ğv*[ေ ǰေp

ေ
ǰ
ေ <


¿ေ Ğေ ƪေ
ƪ[p
Ɣေ ÔĞေl

э Yေ \E[ ,Աģ O - Ö Ա8*
o 5CJ5F5o ԔR[ေ)J@H  o
)&;e o ĸ:ေ ÎЯ*pေ 4*³*(
ေ X
ေ 75&Fo ԋ( g ေ + @C o
ေ ۲À੠…0]ေ ÔȦ(ေ + J @F  Oo ŏ:ေ
ÎЯ*pေÎ
ေ *౞Ԛ %ေ C*]YေC*( *p
ေ
)&;f o ›gေ ๺ ‫ ض‬ᛦ Č
е
Ĝ
*%ေŠAေR<v
[ေǰေƪ
ေ‫ݔ‬౟࿷ ǰ
(
ေв <
]¿
¿ေǰေp͠óó % т ေ ฿ٖᛦ
åσ̋Ա\EƢԱ
 ,ſdåE  Ǝ Ա / F&aH;\o ÔȦ(¿ ေ ,J@H !o
)&@!o ೏ᛦ ६ ࣪ ‫ ئ‬ᛦ l8[5<YေŠŐ
‫](*׽‬ေp
 % Ɣ ေFƲӢLjԱˆÖԱ86
o F72@&o Ôŏʲ(ေ)J= o
)&C o Ÿေ Ɋ  ေ 4ƪ
%ေ Š|ƪ
ေ Ğ
ʍ
(Ğ(ေ ¿<
pĜ ေ ǰေ pĪ
ƫ]ေ
p*

<*pေ vĞ
:ေĞĄ % Ɣ ေ
OǚdԱ ॐᛦ 2 Ö ԱB'
o H,;H)Ho ÔAҦpေ ,J;)  o
)(F o 4  ေa ƪ*
5
 %ေ Š€Ğ

(Ğ(ေɊ

Q:ေǰေ4* Y Ɣ ေ|ƪ
ေÎ
<
ေ
 YေȦǏǰ<èေ
)J@, !o
)&H!o   ေ{  ေனޫေ
ေA z ေ4zေ ҈(%ေ Š|ƪ

ေ5p*ေǰေ*55(ĥ<*(ေ`5]ű:Yтေ E ƦŒԱ
d/ - Ö Ա' / F@@FC5o ÔA<ਜ਼ေ )J7& o
) &H d  o ĸ  ေaေ WĞ5ƪ
ေ
ေ ”gေ ”gေ Ɋ* %ေ Š€Ğ5³5(*(ေ ƌóေ ǰေ Ȩေ *Ȩေ ȨЯ
<[ေ
Ðʍ*(
5ေp]
%Ɣေ฿ٞᛦ
±Ա O'Ա
å, ˆ Ö Ա:6
o )&F ) ) 5o ÔŚ
 zေ V@Jo
+ &J  o Wgေ lေ 4
:%ေ ŠǀĂɌ
(Ąေ ǰေƪƌ͠ေ ေ (ƪ
*(]ေ p]:]
% Ɣ ေ฾ ٔ ᛦ
ӆ  Ա ӇԱ
å,  Ա
:'
o + C;@)C@7o Ôͣvေ W1 / H  Oo Šl]
ေ
ေ v]ေ ǰေ p]ó % Ɣ ေ F*Ա  ƈ  Ö Ա *6
o
;;H;@Ho Ôŏ
(ေ )J)^ !o
) )& o ”ေ ǀ  ေ 4 

<%ေ ŠWȨ*R
ေ p

ေ ]ေ v
Ɔေ Ğေ Ğp [ေ
ҧ
R
ေ pʍ]ေ `RĪ`%ßေ
F*Ա k ƈ - Ö Ա* 8
o37;7Fo Ԕ8
ေ )J) J  ! o
) / ) ! o 4  ေ ܴьR ]Yေ Š4(*
Ĝ*ĂЧ(ေ ǁ8
*ေ  ǰ ေ‫(]ݜ‬RRေ |
( *Ƴ8
ƺٗေ ”ƪေ a*ʍ
[ေ ¡ᛦ
4 Yေ ‚( g % ေ҈Ȩ:ေ ʜ<ű%ေ c;Jo
4 ) 5 ! o ŋ?ϩ Č
pR*<Yေ Š|8p
ေ
ေ ǰေ ƪ*pƪေ
Ă෠(*
(‫م‬ՖƫĿ àေ Ɋʍ
(Q
*pေ ǰေ 8póĥ
ေ
ʍ
ေ
ေ
Ë ေǰေ<
v
*pေ*YƔေ Oǚd Ա
ҳƳԱƁd#Ա2ė ƦԱ2ӡ/O  Ǝ Ա86
) J + L-
എǞླૺጌᛦ ÔȦ(gေ )J)7  Po ‚Ŀေ͘*p
ĨĂЧĪʍ
ေ]ó%ေ>n>Ÿ- ! Ա)J7@)J@<!o
) )M. !o Ɋေ ”ေ |%ေ Š‹8¿p
*pေ ǰေ p
%Ɣေ ºCԱЈѮssǢԱ F˲Τ-ÖԱ 6B
) 5RH/ 5;7o ÔȦ(ေ )@@ Oo
ŠǀƳ8*³*`<*8ေ `
е

ေ p
ေ
ေ :͠Ȩေ v
Ɔေ ]ေ `
QƌĨ8ေ ¿55<
ĥ8
%сေ
>n>Ÿ  Ǝ Ա6G
) )&J))5&o ԋ(ေ )J@C o
) );  o W  ေ {  ေ A^ေ Î
:
[%ေ ŠAေ 8*v
<
ေ 

p5ေ ǰ<ေ ̣
]*pေ pʍ
:
<5ေ ]ေ *ʍ*( ] ƺ Ɣ ေ
ſ ǩƳ Ա dŸ Ö Ա:A
)5&@)5&Co
)J@7  o
) ) @ ! o Aေ lgေ î(ƪQ*ေ
ေ {  ေCေ a
``
<%ေ Š Ţ<ߐ¿
(*v
ေŇɚໞÊ

%ٜေ ”ƪ ေa*ʍ
[ေ ¡ᛦ
4 %ေ ‚(  % ေ‫
ݚ‬:ေ ·<Q%ေ + J@7 o
) ) C ! o j ( æ lေ €<p5%ေ Š4*

ေp
 % Ɣ ေƠ>E>sԱ  ,Ա ƈ Ö ԱI
o 5Fo Ԕ
ေ );C Qo 
5ေ ]ေ
\§ԱdŸ  Ö Ա6 G
7;F7;H  7C;o 7C^7F&o ԔR[ေ )J;C  o
) + F ! o Wေ aေ ŏ
p̡
%ေ ŠC*pေR`
ေ¿
ေǰ<ေ:

ေp
 ড Ɣ ေ4Ą
ǰ໼‫ؖ‬ေÑ*v5<Ŕ
*[ေŸ5(*(ေČ]`]*
%ေRR`Ğƪ
ေ<
ေေ:Ĭေ*ေ
ေ‫׭‬R`
ေ|
( ࿞
*ƳR
ေČ]`
<[ƺေ
)J@; o
) ) H " o a  ေ {  ေਫ(›*p ¿ Yေ Š<

ေ
ေ *ĥ

ေ p]ေ :
<ေ ƪ

%Ɣေ ୿ʄ

Ğ
óေ ¥ᛦ
Ӫʄ5  ƺေ :'
o J;JCo ԋ(ေ )@;  ! o
) )J  o ”ေA g ေ”ʍ[%ေ ŠAv

p
ေ ǰ ေ(

*(ေÐ  ေ
ʍ
(ေ 8`
%Ɣေ҈
*
ေŸ
(<Ğ(ေ
Îǰ
<
(
%ေ Î *(
p%ေ Ȧ(¿ g ေJ  o )J@F!o
)5&o ࣲ ‫ ط‬ᛦ 
[(
ေ
ေ”ေ4
Η Rɲ5%ေŠl
ေ
ေ
ေR`Ð(]ေ*ေ
¿
ʍR<pĹ(
ʍေ 
]Ð Ɣေ
ÔÐေǁ<5( Ē Yေ Ó OO Ա [Ա)ɵO,[>) ÖԱI
)7))7@o ÔC]ေ )J@;  o
) 5 )  o ”gေ 4 g ေ€
Yေ” %ေ Š4

*pေ*(
ေေp
ေေ

ေ¿  ေǰေ
ေv
(R8Ĩ*p ေǧ * % ٜ ေ
 ƈ-Ա
,>ԱƁd[)!Ա*8
o 5CDE5CFo ԔR
ေ 0J;5  !o

) X kX
Ա X Ա — 5
Ա ™kԱ ̍Ա DZʊ
Ա

TTZေ lေ tĤsေsœေmေ Щ٣•sQQ
rေ d̜
çေ/Ǹ•Ǹ/ေ}ေ/Ĥsေ•ŨœေvsvǸ/ r o ေƠv:§:Ñ@Ա Þ%,v«Աk%ƈ G ¡ Ա

̈́ ij T1GbTDzေ Fṯေ ̙Ή±©¶ ေ
HT ေ A Z ေĮ Z ေ€sǸေsœေ ͍Zေ Į  ေ4sç•- rေ dAေ
ေေv
•[ေ /}-ေ s//
/ေsœေ/ĤǸေ }ေ’
•ေ


-•sေsSSs’/ Y o ေ ºCԱ
,Ա<R@@Ա ˚%,vR C ! Աbx BԱ XZေ±©qေ±@૔ࣣɁ ֠ေF͍R
ေ
G±1¶ေေ
T֩  ေ AZေ Į Z ေ€s
ေsœေ ͍^ေ Į  ေ4sç•’ Yေ di
ေ œ
vǸSĤ
ေ}ေ/Ĥ
ေ vǶ•[ေ /}ေ s//
/r࢈ေ
º « Ա
, CԱ<§@@Ա%,vR « ! Աbb BԱ XေTrေ©@˜@ေF͍R
ေ˥ѽ֩G¶ Zေ
ñ TZေ aZေ Į/Sķ qေ dÞေ /Ĥ
ေ 
çေ s//ေ /œ
•/ေ }•ေ vsRRĤေ 
ȄÐƏR
/ٛေ Fေ ŠǸ•Ĥs ¶ è ေ
™%:•%Oîǜѐv« ¡ Աz z B Ա … b …©ေ FG ¶  ေ
ñ T©Zေ ·RQssေ ÷Q
œsrေ d4
sေ Ĥ
- œ/ေ }ေ vsRRĤေ Rµ
/ေ ç- ေ ç OĤķÐေ s// Y o ေ
Fေ͍sSsǸ/Ǹ¶rေº C Ա%OŠCԱ
ɭCԱºÑ!Աz‘BԱX CԱ ©©rေ1…b1±ေ F૓˾ ¶  ေ
ñ T@ Z ေ { Z ေ›s/qေd4œ
•ေs//Ǹ/oေFေl
•Ĥs¶qေ
ÑO%ɮv@RC¡Աut BԱ̜Zေrေ © b ©ေFG ¶ ^ေ
T1 Z ေ tေ ŏsĶrေ d4œ
•ေs//
/ေ}ေ Ǝçေ
uSs/oေ FÐေ m•
 ¶rေ4ÐÐsǸ/ေœ༖ေ̈́ijေ THɝ
T±ေ FtRေ G ¶ ̑ေ Ž̈́ijေ @b @1ေ F4
Sေ G ¶  ေ
TGZေ { Z ေ›s/Ðrေd4œ
•ေs//
/oေ Fေl
•Ĥs¶੘ေ
ÑO%,v@RC¡Ա̈́t ijေ ଒ᛦ༣ ‫ ٷ‬ᛦGqေTñTbT©ေFGνϴ¶ Òေ
Xေ ±rေ Gb ± ʂ ေ̲ေ ©q 1Gɹ GTေ
 …Zေ tZေေ{ss/
ေsœေ¡ေĮZေ {cqေd4
sေǸ ƏR
/ေ}•ေĤÐs-R•
ေRµ
/ r o ေ%aǩƳ «Ա
( /% ! Աtt BԱ ©˜1ေ FXvေ G¶ Zေ
H  Zေ Wေ {  ေ ĸsrေ d4œ
•ေ s//ေ /
s qoေ ºCԱ XŠCԱ ǜaʍň CԱ
,  ! Աb‘BԱ …GbHTေ FCs•ေ
G©¶ ေ
 TZေ ͍ Z ေŠsRSေsœေtေ lေ mေ ŏÐçs³rေdޕS
•Ð
/ေ}ေçO’ǸĤS
•ŨR•
ေ/œ
•ေs//Ǹ/ Y o ေ
%O'CԱ ªEE  ¡ Աbt BԱ ±@bေ Fmķµ Zေ G@¶ ေ
ñ Zေ ͍Zေ ¯
•Ĥ

• rေ di
ေ ĤSR/
ေ µ•
sQœçေ /’•
p’ ေ }ေ [•Ǹuေ s// qoေ FvĊ@:ǽ˄!Ա ŕǫijေ
H b T©ေ FHG±¶ ေ
ñ ֩ေ ᐴ‫ز‬ᛦ ¯Ǹ•Ĥ

•rေ di
ေ

•ေ /¿•
’ ေ}ေ ps//
/ေç ေœÐФ}Ǹ•Ǹေ /œRĤေ-Ƕ/ qoေ
υ φ ֦

:nƀŸ « ! Աŕŕ ijေ T±@ɝT͇ຜေ FHG©¶ ေ

̈ ေˆ ᐴ ‫ س‬ᛦ‫
׮‬ϙĤǸǸ•rေ di ķေ ǸǸ•sေœRĄ Ljေ}ေps//
/ေsေ Ð ေ‫ބ‬ƍǸœေ/•
p-Ȅ / r o ေ૒ଡ଼ຝมภඅZqေ
χ֦

 / T@ T©1ေ FG©¶ Zေ

́ Ѳϴ©Zေ ͍Zေ¯
•Ĥ

• rေdÞေÛȄ
ေ•
sေµ
’ç

ေေœRÛÐvÛ[ေsœေµ•ʲsQœçေ/•ǸpೳԚေ
}ေs// r o ေ:n:ŸC _ Աŕŕ ij T©GͱT@1ေ FñG©¶ Zေ
ୱ@ေ V  ေXေ V
őQ
•qေ ¡ေ { z ေCsཽ/ေ sœေ Éေ {Zေ ȧs s•s/ rေ d ŏ
vǸSĤ
ေ}ေ cRœေ
s//ေ µRµ/ေ }•ေ W¡AOT¡ǦTTေ •ေ S̱Ɔ
ေ -Rµ
/roေ SsS
•ေ S•
/
ĥǶœေ s’ေ ’
ေ
G@ေ¡}
•

ေေĮ
’•ေ ŏ
vǶ/qေas/ Ð-qေ ౰‫ش‬ᛦ ¡ Z ေ
 1ေ ʙ  ေ4 z ေ ms/
qေ tေ ŏ ü ေð
s•/ေs  œ ေaZေ W  ေXေ • v
•qေ dǤçOĤ
ߢ ေ •sေ
s//
/ေ çĥ ေ  ေ Ĥ
-ေ ԕRœ
/ေ µ
çေ±……Ħ¡ r o ေ ºԱ ‰ ŠCԱ ƤƥOǩƳ ˆԱ
ɯ C ! Ա  / ±…ေ
Fʙ
SZေ GG¶ ေ