Вы находитесь на странице: 1из 1075




K.C.U., D.Sc., F.R.S.







arc due.
ivo workers to whoso efforts thoy
To M
out what are tho latent niothoclB
,wn aboutiho law, of Friction,
of Oalorimotry, whul exactly^

how far ha, Iho theory of llu, Htomn J^ino

01 n
tho on which mcilhoila of aceiimlo g-uigiiifi
mncod, what are principle* of a l h
factor, which conio into tho luy-oul

ct ^ It
dote nmnation of tho numy
he design of a Dynamo, or
nn" a lontj BO-ireli in Libraries and,
llLon.otliodH.of Pyromctry

nob infroqnoiitly, a futile jmirnoy

,tro based,

t.. tmmo

ho found
wiHliocl-fov information may
ic" whoro'it ii liopo.1 tho W McU o,ta
of Optical IiiHtnimo ntB,

The Scionco of Aor.n.utic., tho Doeigii thu J.IWH

TeloB C opo or MwroacopOH,
tho Construction of ClorM
Mimip and Acoustics uro all baaed on Physics.
indeed, w,t ,,,uu tta.
T ,lL,r who U co,,ce,,ed with .hone
that lin boon dono-that
would bo

know, not perhaps all

l,or B nl)ioets nu, a t
to .nfonna-
find tho latnt and most
u heavy a task-but where ho may

, from
~ ^^'-\
nanJel of so.no of Iho principal eontdbnUms
tho help of those ,uot con.po.ont
M-tor h,.

wnto ,m tub
", L ato h, wiring
Ho aIHO , o t ,1
been produced
r h< the Dictionary could not havo

Cou,,oi. S have allowod uo ,,f ., ,v

H ,,nb r
of Sclcntlfl. fk-doto ^l.oHO
U1H -from theiv 7Vi/ to be froely ,nalo. Among
ll 'l'

Mlm,,..nl l.nco.

tho Institution
ionert in particular tho Koyal Socioty,

and the Institution of Electrical Engineers. The same help has bcoti raulily
itllbrdcd by it number of Pub] labors.
It la clear that, with so largo ;i
range of subjects, any individual worker will,
probably, bo concerned mainly with 0110 branch, ami, tliis in viinv, tlic

volumes have been aminged, us far as jinsaiblc, in snbjocts. To obtjiin informa-

tion as to the latest advances of Applied Electricity it not bo noeossuvy to


purcliaso the aections of the Dictionary dealing

witli Aorouantins or Mutcsorology,
Tiio iirrangcuicnt in ouch volume is alphabetical, but, at the saimi time, it lias
been thought best to deal with each main subject for example, the Thevmo-

(.lyuamies of the Steam Engine in a continuous article ;

references are given,
each in its own alphabetical position, to the headings of tho various fictitious of

ui\ article and to the more important subjects which it includes,

T1 .

n A ir p HI Iu9t.C.K,
]3|, w tic Constant*,

Kincinntics of

SlA^aiu TSngiM, Iteciin-oculiiiB. .

i? r; ^ M tnfit,O.K, M.I.McchJ'1, A.M.l.N-A.,

Uii^ of-
Enine ail Prime Movers

O.B-E, H.I.A.K .

EVASB, Aumnav T.,

Poti-ol Ifingine,


J. H., A.M.Imt.C.15.,

Mechanical Power*.

Pressure, Measurement

^ Hoiuwro, SC.D, 1?.B,S.

'JToiU'icr's Scries.


Harmonic Motion.
Stroam-Uno Motion.

T,AHDON, J. W., M.A. .

Structures, Strength

LEVY, HYMAN, M.A., D.SO., F.11.S.E. \''
Dynamical Similarity.
u K
- -





Elasticity, Theory

STAHTOH, T. IS, 0.13.13., US,,, Hl.ln B t,O.K, KH.S


D.Sc., M.Insl,C..IC, KH.H., and

Steam Turbine, Physics of thu.

BwTi'iinFiKM), W. J. A., M.A., V.I.O. .

Fuel Calorimetry.
awl -

Sir DuaAMi, M.tn*t.O.E., M.I.Mech.E., 17.H.S., )

A., M.lust.C.K,
M.Inst.A.E. J
BURLS, tl.

of Internal Ciimlmslion.
Engines, ThurmoilyimmicB
Internal Combustion.
Engines, Some Typical
A. i,
. .
aild B.,
L., I'h.D-,
Ph.D.,| ll
U.S.A. J -

Gcopliysk-al Institute, Washington,

of Absolute, Scalu of.
Temperature, UeaJiaation
I'MiS ^. A.

Liquefaction of
Steam Engine, Theory of,

Tliennotly nan tics.

V.R&. '' ; - ll -


Heat, Jleohaiiical li(juivale]it



.....' ri

Calorimetry, Electrical

Calorimelry, Method of

Caloilmetry, Methods liaswl

on Ohangu oE State,.

Calorinnitry, Quantum Theory,

Latent Heat.

Pyrometry, Optical.
Tyrometry, Total Radiation.
Hoaistaii co


WlLT.fAM 11 ATM, JLA., I'Ml.H. -


JOHN L., 1>.M(!., KIiiBt.l . .

LAM lloitACK, S-M-.O.,

I'Ml.S. .

Jouduiaion ol' : ,MaUimiuit.ii:al Tln:ory.



GHHUH, Sptioifm
SLA. ...

]To,'U; ol

Honoii'iHi.iJ, K II., .H.A., H.H(5.

1 1
Conduction ol'.

lldtvt,, OunvuuliiiH of.



L'UMi'S. Sets Air -


J[ a pir[i!i;tly
rovorsiblu hnat ciHiinn UUUJB -

An no DYNAMIC TACiiKOMETiiR Jni' measuring


in iv quantity of hunt Qi a^tompuratum T, 1

unit timo by
Q at timipiii'atum T a tlmn ninnlior of nivolntitiiiH iitu
and rojootH a
iiH'aiiH of nir iirosauro ilifforoncos.
(). l l'K l
-Q a /l'a provided tins tmniii'iutiireH
Mutora," Vol. TIT.
T, and T a am moamired on Hid alwuhitK 5 01),
AKUU--KNUINI.;, Tun llor^s-KiiYon EAGLE."
Botilit. llcm^ thn ratio "
tliat Hdaln Kiso Potrol HiiKiiw, 'J-)c Wator-coiiLod,"
of two limi|imiturcH on is i!(|tial

lakuii In to th --
to tliti ratio of Lliu lu-nt
, v
Jiout mjuutuil by any porfticlly

ongiuo working
bolwcou tlioao tmnpnraiiun^. piinipH," g (S).
" 17 )' ( 2 -) Allt,(JUNHTITIIKNTH Of, SKrAKATHI) 1IY IntAO-
Sou TliormodyiiainitiH," 88 ( '
" of .InLninal TfONAF, :i)WTILLATI()N. Suo GaSOS, LlllUO-
.Kiitdnoii, 'I'liKnnodynainiiiH
faction of," (ii).
" uscwl as
<'N C)AH Aiu, JNDKX OF HuniAtiriON m',
tomiionitiiro in

AND "WOIIK" HUAliKH. HlIU HitiMindnry Hfcnndiu'd "

itinu ftlicivo IH)U (J. Hco Tompomturo,
dymuniuH," ('t).
KtnUiHatioii of Alwolutu Sualii of," 0>1) (
ll -)-


Am, KfKUiKict JlMATov:
DHitCI'H," (2). Ron "OasM, bjicoiUo
At liigli U'iniwrotwro.

Iloalof, at JMKhToiniwratiiPOH.
Kl-JAIHNH )!.'

At 00 fJ. mid various prcmnwH, tabiilotwi

" Total
BIKTBR8. Hoc Pyromutry, and Jacob.
vftluoa (il)tainod by Holliorn
Uon,"(]0). " 3'Mcotricat Mcthotla of,
fioo CliUorimoU-y,
(10), TnWoX.
Variation wltli prcHdiiro over tho rango
Movalilo. Hno ibid, (!)). cldtornunod by
1 to 1200 nlmoaiilioroH,
Ati(!i3i,iiiiATioN IMAIIHH. Kco
Jlolljom and Jnooli. Sco i&W. (US).
of iVachmovy," (4)(iv.). IlWATfl OBt
hco Alii (HBB FKOM CO,), Bl'KCIIfia Sohcol nnd
tobulatwl vniiicB obliaincd by
. " Jl-loctnoal
IloiiKO. Son Oalorimotry,
AIJIAIIATIO CiiAiron. A n\\tmw ]n tl10
a body otM'Hwl out l

aiul iiroBBuru of
a way that no hunt ia allowod
ibly i siiuli
Sue aim (lotorminoil at mom and liiw toniporuturw
to pafin to or from tho body. molhoil,
" bv tho cnntinmiuR How oloofcnun lt
Tliormoilynaniios," (JO), (118). Soo Cnlcirunotiy,
CTAH. ]y ftcthool and Honso.
" of lamiti'iwil Methods of," (IT)).
Sco EnginoH, Tlint-motlynnniioH

Comliiiutinn," (20);
" Motors fin aimiHiiromont of Stoani,
A ,, Vol. HI.
EXI'AKSUIN 0V .In.UlI). <r>) t


(38). "
THOTIimiMAL (JltANUKS. K(H1 Hwi HiiEriKBratlon," (4).
AH!) " IT. 5 I' 1 ")-
nf Tntuniiil Atn-Lirp 1'usiP. Hwi irydranlicii,'
, TlioimodyiiaiiiiiiH
Am MBTKUS. Sno "C ftl-B8
Wl Air
(3) J
vi. nr.
T n TT. Tn = absolute temperature of atmo-
AIR-PUMPS sphoro, UP.V. aiid'H.P.V.
m = mass of gas.
S volumetric speed.
An air- or gna-iiurap IB a device whereby gas W = work.
la from a low - prMSiiw vcMol
transfm-red to power.
vessel V.)-
(LP.V.) to" a high-pressure
(11.1. p~ density.
The term vessel includes tho froo atmo- ij
= viscosity,
" "
Hphoro, and tho
term gas includes vapours. e = friction cooflieiont.
It is assumed, unless tho contrary
is stated, 15 = 'JSmecii. = moohaiiical ofTiciomsy

that the L.P.V. and H.P.V.

are at tho BAIUO
E,, (ll
= volumetric oflicionoy.

arc separated by a
If the L.P.V. mid H.P.V. (3) WOHKINCl
gns-tight partition,
and if tho gas is not a may be distinguished either according
to then-
diminish tho or according to tho
saturated vapour, .tho pump will working characteristics
pressure In tho L.P.V.
and increase it HI tho
principleo on
which tho action depends. Of
H.P.V. it will act at tho
same time as a tho characteristics tho following aro
In practice tu
compressor and as
nn evaouator. tho most important of thoso applicable
one of tho two vessels is almost always pumps ol all types :

and variations bo worked

tained at atmospheric pressure, Range of Pressure. ft any pump
of pressure in tins other vessel
alono arc im- there will
continuously between closed vessels,
If this condition is fulfilled, a
com- in them constant
portant, ultimately Lo established
lie defined as R ])ump
of which tho range of pressure
proBsor may pressures, p n , pi?. By
an is meant either (a) tho ratio j>ii /V,
()l ( 6 ) tll(1

tho L.P.V." is at atmosphorio pressure,

difference jj,,' -
as one of which the H.P.V. is at tho nioro
ovacuator (<0 V-
hero ho termed
atmospheric pressure. important quantity and will
Tho L.P.V. ami H.P.V. arc aclclom com- the
range," denoted by
K i for it oftou
in laboratory evacu- of tho absolute
pletely separated, oxcopt approximately independent
ates there is a continual stream o
gas values pn, P*>- But it is never completely
pumps havo a minimum
from ono to the other. If tho energy independent; for all
this stream ia comparable fcolow which they willnot reduce pr. what-
quired to produce
with tho wholo work done on tho gas, ^tho ever is the value, above this limit, of ?>,
pump may be termed a blower," or, if it is and all havo a maximum pa, though it
of one constructional typo, n, fan." The bo determined only by mechanical
distinction between pumps blowers,
though formally indefinite, is perfectly Tho range of a pump of any given typo
Blowers are usually, but not or mwv
in practice. may be increased by working "two
always, compressors, producing pressures pumps in series to form a compomtci
In blowers there of one being tho II.I'.V.ol
fciwtor than atmospheric. pump, the L.P.V.
can be no single and definite p n or p r,, capable tho next. In all important eases, tho iunin
of general scientific definition ; but there ia or
of tho composite pump is approximately
of places along tho stream of tho ranges of tho
usually some pair exactly tho product
of gas passing through tho blower at which can
components. But a composite pump
it is obviously convenient to measure pu also ho built up of components of difforont
pL . These places may bo regarded for our types ; no general
statement can ho miulo
purpose ns constituting the fl.P.V. and L.P.V. about tho relation between tho range <vf
Suffixes L and II denote com-
(2) NOTATION. such a composite pump and thoso of its
to tho L.P.V. or H.P.V.
quantities retort-ing pononts.
Many of tho formulae given will still ho
true Tho snood is
if tlics suffixes L and H
are interchanged j this the rate at which gas is transferred from
foaburo is indicated by writing before them tho L.P.V. to tho H.P.V, The amount uF

{" L or H "). gas is usually estimated by

its volume at blio

pressure of tho L.P.V.,- whether

the pump IH
P, yt, j)n =prossuro, Tho H|i<n"l
= initial pressure (tho same for a compressor or an evacuator.
so estimated is called tho volnmotrii
L.P.V. and H.P.V.).
= final preasiu-es. speed," S, and is expressed in volumo \it
2'j,, j)n
unit time.
Measurements of S aro usually
n=atmosphoric pressure.
P= vapour pressure. observations of tho change of p\ t \
in an ovacualor, tl.
V Vn = volumes of L.P.V.
and H.P.V. pressor or of ps,
or L.P.V, being completely closci;
iii,> MH= maximum and minimum volumes
of cylindor." during tho measurement is email I
to an evaouator ov compressor
bo regarded as
pi,, tho gas may
or (3) .applied
with p becomes
perfect, Consequently for nn ovaouotor
(LorH) W^^Vn-^ -

In a compressor, pi, is constant and equal to

II. Therefore If II - smilll > (

s= . . . .
w (LorH)
in S is
In blowers tho volume involved If pjl ia small, as in a high ovacnaUir, it-

It ia conveniently
usually estimated at p K
becomes ,_,
mensural by some typo of flow-motor placed
If tho pressure
in tho outlet or itdet pipe.
dono by tlio
at fcho point whore the meter
is placed differs In aomo text-books, tho work
corroetiou must ho of account; the term
considerably from Pa. a atmosphere is left out
of tho motor. in II is omitted from (3),
and tho second and
applied to tho readings from (B). J*ul
S is usually a function of pu and pi third terms in tho bracket
o.- on the atmo-
hut is always dono by
aa well as of the nature of tho pump since

or evacuating, the
there arc important exceptions.
Tho range sphere in compressing
of reckoned without those terms
K or tho maximum difference
00 pressure efficiencies
Boom to have no practical significance.
is usually
that s=o. . ,
In Unworn tlio useful effect
, ,.
Several kinds ot estimated by tho volume of gas produced _at
g (5) THE
to force
MO recognised as applicable) to
efficiencies a eivon pressure. The work required
into a
pumps ami blowers ; of the tho mechanical a mass o( Rtw from the atmosphere
IK mmntfimed
ratio of tho useful work vessel in which, tlio pressure
pmmcnoy, or tho fchfl
ia alone constant and ei|ual to p n 0>y iiwrPHBinf^
ilono to the total work expended,
w Riven.
applicable to
all types, Botii terms of the volumo of tho vessel as the gas enters)
to rid them of /m
ratio need further definition y _.,-

ambiguity. Tho work expended ia usually

taken to mean either () tho work expended V is the volume of tho pas
at pressure
com- where
on tho gas in (jiving to it energy, if" 8 is the volnmolno
or thermal, or (6) tho worlc MI,. Consequently,
pressivo, kinetic, and w the work dono per second,
to tho mechanism of which the aiiced,
supplied . (0)
nump consists, including that
lost in fruition a^n-lLJS. .

of wlid or liquid parts.

Tito olhoionoy of from a blower
Tlio stream gas issuing
reckoned with (a) to oflen termed tho
It tho worlc ex-
possesses kinetic energy.
" " that reckoned with (b) >" *<> '
gas efficiency ;

pended in giving to it this cncirgy

" ovor-all " numb ho added
the efficiency. included as useful work, there
In pumps, where the L.P.V. and Il.P.V.
term p v = W>
taken within tlio braokot in (0) tho
are separate, the useful work is always It is often impossible practically
to convert
to be that required to transfer
tho gas that of any ot>or
this klnotto energy into energy
IIRH actually passed from
tho former to tlio which m the
form without stopping tho flow
latter. This work will be least if the trans- main of tho blowor accordingly t!io

if tho L.P.V.
ference is effected rovermbly ;
it In called, reckoned from
total efficiently, as
and H.P.V. are at the sumo temperature,
tho relation , ,

revorstblo transference must

he isothermal,
and any change of temperature during
tho expenditure of more Hnt it may ho nntod
involves ia often miHleadmg.
process to reriuco
work. If the transference is rovcrsihlo that ideally it in always possibl.!
to transfer a anil thus to
isothermal, tho work required p to Koro without ehanp;inp; S,
con- !<'
mass of gas between tho atmosphere at "'convert velocity into prt'HSiirn ;
the orose section of tho
Btream is mtwlo
stant pressure II and a closed vessel,
in which ia changed by tho transference an infinitely smnll v will give
infinitely largo,
from H to pt ,
is given by n finite S,
In addition to tho mechanical cJliciomiy,

them ia recognised for many pumps a quantity

known as the volumetric offlutonoy. Hut
this quantity cannot be defined ((onorally
whero V = fij) is t"
isothermal wharnotorlBtio
ho discumird in
the elmcd all types of puni]i, it will
of the mn'sa of tho gas occupying - '"" 1

If tlio giw IB perfect, oonneotion with thoso to whioli it

vessel at tho pressure p,.

(II). Tho remaining working i;luim<itemtics but will ho conveniently noticed briefly at tho
ooiniiiiMi to all pumps aro Iowa ciipahle of preeiao end of this article.
nuMSHrtmiont j but they nm mnw the less
A. AllKOSTATff! l.'HML'fi
important. They includo simplicity and tson-
voiuonec!,lii'Ht coat and cost of maintenance, (8) HANOI-: AND (-jr-KHi). Tho working part
adaptation to availo-Wo sources of power, and is always a vosstit (U) of variable volume, u,
NO on. WhoH
types of ]HJJii{)
jir- an Ui.s
ini!tihlo, usually those chmuo tomtits
it ia
(miH!tecI to tho L.P.V, when it
mthor measurable) efliiupnoy which
tliau uny volumes 19 a maximum i, ;
tlotormmo tho elioico. They will bo noticed distsonncoted from tlio L.I'.V. and the
in <j(i(irifl(!ti(Jii with [KH'tioiiliir tyjjejj. volume decreased to tho minimum ui\ ;

(V) .l.'inNcii'Li'is OF AOT.ION. .I'Viv tho

(3) connected to the H.J'.V. ;
t'lotiuled consideration of tho various types of and tho
(<i) diBComiocted from tho H.l'.V,
pump, it in KIOTO uonvcnicrUj to adopt a
voluino increased to.[ |P

olfiHsifumtion fjasfid upon tho principles under-

lying tlio notion, Hero pumps fall into thi'co This ideal oyclo is never attained in praetieo
great but forms tlio basis of any calculations.
Kvon if tho ideal eyob were attainod, tho
A. Aerostatic.
general formulae giving tho K*]lioti botiwooii
B. Aerodynamic.
pu, pi> fla after a number of cycles n would
0. Molecular or Hi&h-v;umum, as-
ho nxtremoly complicated. -lint if it is

In mi ei/fl(i'fi pump this transforeniio of sumed that the (fiis

is perfect, and that tho
gust is o (Touted l>y forties that arc at any instant transfereniici in isothermal, tlio relation liotwwm
in aliii-itml or/ nil iljri urn. Kt>!' uny }w'tieu]ar (?}([) and the. vivluea of pn after
pump pressure ia independent of
bins raiif-o of tho >ith and {-H-i-])th cycles, ia

thn n|i(!dd working witliin wide limits

i> tlio ;

)niiiip nan Iio worked infinitely slowly without (L<irH) (./(,,^^^1^-Xul^iIli. (1)
Mi''' "n
?'ftny or of oiTioionny, (This atdtomenfc
ia not Htriclly triio wlion tho viscosity of n In a compressor or evacuator wo havo from
hdivitiafcing liriuid its used to Becnrc gas- (10)
fuiuh pumps am dynamic, lint not
In ;tll {micticjil oxftinplcs filto ' i
' w "> fi"n1

statical (orces aro those duo to comjirftssion,

lint thoso duo to ohangn of tornporafcuro might (11) IH trim wliatovor tho ratio of n to tho

wiH!iMvaI)ly h used.
Vs ; if it is small and if N, tlio niiiufior of

.lu cut ae.fO(ltjRttmie- piutiji tlio forces <m tho Rydi'.s [U' unit of time, in )/ujo, (11)

(,'iw aw iVyimmiuii), and vary witli tlio motion

u tho purU of tho pump; they coaso whon
tho apnfttl nf working heciomcs itiliiiitoly sinaH,
HCI tliat tlio range and sjrertrl (F tlio pump (12)
vanish Loyotlior. Those dynamical forces Proiw (12) mid (1) wo obtain for Uio volumotrio
from or of the
tho inortiti visooaity spend of an ovaouator

'('(IP (Itstinclion IwUvooii fclio two classes

unn bo oxiH'OSHod liws furiiiftlly, hut noiJiftpH
inoro oltinvly, hy Haying that in thn first oluaa, and for thnli of a
" "
bub not in tlio Ncuond, liho acition la jmnitivo
in tho ong.incoi'iiiK HIUIKH <ir that, wliilo ;
it in

impossible to blow fcliroiigh a of pump ilio

firat class, it in possible to blow through <mo

Tlio lutigo of presstii'o is
given Ity iS 0.
of the sooond.
In hotU thoBO tifasaos tiio forces aro umih RH Consoqiiontly
nra asawiatol ivitli a oontiiniious modiiiiH.
" " (LorU) *?*~l\
. .
In tho HuMl class tins notion is duo to forcca
ajiproeiixblo on tho motconldi but not on tho

The mngo, measured by the ratio of tlm

innltu scute.1
Tho clasH would properly bo
" preshnvcH, i independent of tho initial preHHiiro
termed mdositlfir" but sinco that town ami of Clio volumes of the L..P.V. and II.C.V.
has hcori ttjiproprmtod to a jinrtioutar member TJiflso relations become voiy nimpln ivlioti
of it, tho 1m soioiUifio expression high- HII, tho voluino of tho
dead spaco," in wint.
vasunm pum])3 will ho used, Then Iwoomo"
In ntWition t thcso tlirco classes of pump,
aro some methods of ovivmiation which vl- = 1 + rt (eoinprossor), .
ly satisfy tlio doflnilion of pumping, 11 Vii

? 'f'
IE the H.P.V. or .LP.V. is not at atmospheric

(17) temperature, (20)-(23) must be corrected by
tho substitution of Vji/T H V],/T(., KL/TO for ,

H Nit (compressor), . .
Vn, Vr,, ?(r :


(evacuator). (10)
vantage which aei'ostatic pumps possess over
other types lies in the great range which can
the volumetric speed of such a pump
Thus bo obtained with them. They are, therefore,
woidd bo independent f tho pressure against well suited for the production of very high
which it worked ;
its ningo would be infinite.
or very low pressures in a single operation ;

(()) but extreme pressures can be obtained with
actual pump ta tho ideal cycle performed :

other types combined into composite pumps.

the yield is always less than that given by
than that given They are in general less well suited for tho
(lO)-'(H), and it fortiori less transference of largo volumes of gas under
by (l")'(l") tm assumption of no (load
moderate differences of pressure, although
space. The deficiency is due to incomplete sonic types (Ac, d) are used for this purpose.
" " " "
connection and disconnection of the
Their disadvantage is that they cannot ex-
L.P.V. and H.P.V. with U and with each haust vapours satisfactorily, especially when
other, i.e. to leakage and to a
failure to
designed for a large range, for the vapours
establish pressure equilibrium. condense in U when its volume is reduced
Tho comparison of an actual witli an ideal
" volumetric and do not pass readily into the H.P.V. ;

is made in terms of tho

pump when tho volume is increased again, they
wliioh may bo roughly
efficiency" (K v ,,,.), evaporate once more, return to tho L.P.V.,
defined 'as the ratio of the number of cycles and keep p tt permanently at or above the
in which an ideal pump would produce
vapour pressure of the substance. Permanent
the number of cycles in which
given efl'ect to gas mixed with tho vapour
is removed very
tho actual pump produces the same effect. in the L.P.V.
slmvly after its partial pressure
Tho ideal ia assumed to have the same
pump has fallen to that of the vapour.
actual pump
"i. (<!.;/, cylinder volume) as tho ;
The various typos of aerostatic pump are
'also usually assumed to have
no dead of and U
distinguished by the construction
it is

space .
This last assumption is not tho means adopted for connecting and dis-
of tho dead space can
necessary, for tho often t connecting it with the L.P.V. and
be readily calculated if the pump is otherwise The following sub-classes include conveniently
ideal ; but in the pumps for which the all tho important typos :

of volumetric efficiency is most

conception Aa. Solid Piston Pumps.
is always made as small as
important, KH
and its magnitude is important in
Aft. Liquid Piston Pumps.
possible, Ac. Moxiblo Container Pumps.
of tho design.
judging the excellence
" Ad. Rotary Aerostatic Pumps.
by which H vill is estimated
The efl'ect .

must bo defined. It is usually either (I) the

attainment of a given j>ij\l or p a {n, or (2) A. Solid Piston Pumps
of gas
the transference of a given quantity (11) Tim VON
with a llxed j>i, r pn. In either case, J3 vl ,i.
the oldest typo of gas-pump, and its invention
is a function of ?i|, or pn, and
the value of this von Guericko
is generally attributed to Otto
l''or a given pi, or from the
pressure must
he staled.
(1072); it was probably developed
similar water-pump. It has still a wider
other type, being used
(])'und"ror(2). sphere of use than any
If (1) in adopted, and if n is the number ot
for tho attainment of from 1000 pressures
cycles in whioli
the actual pump establishes the atmos. to 10- mni., and for volumetric

then from (1(1) and (17) cubic feet to a few cubic

assigned pnjtt or j^/ll, speeds from many
millimetres per minute. It is equally familiar
. - (20) in in delicate laboratory
heavy engineering,
in everyday
work, and, as tho tyro pump,
of tho type
life. Broadly, the advantages
are a great range of pressure and great
mechanical strength the disadvantages,
adopted, and if N
is the number j
If (2) is
Immsness and mechanical inefficiency. It
of cycles per unit time required for a given
wo have from (18) (10) m unrivalled for compressors, and
volumetric speed H, for
for small portable laboratory ovaeuators j
ail other purposes it can be replaced by
other types. However, it is still widely used
oven for blowers, the
purpose for which its

(evacuate). .
(28) with other types
disadvantages as compared

vork done on tho gns may bo measured, for

arc most innrliocl. Its survival is probably
iho determination of the gay efficiency, by
due partly to its long history and l:o its re-
ndicator diagrams taken from the cylinders.
Homblftnco to the reciprocating steam -engine,
have These oflidciicies arc determined largely by
ut which tho constructional problems ono
the completeness of the cooling, which is
been studied so completely.
is familiar to nil. if the most' important features of these
Tho principle of the pump determined by leakage
The vessel U is a cylinder
in which moves a. jumps ; they are also
made to the L.P.V. and ind by throttling at the valves.
piston. Connection is 2 shows a largo-scale two-stago
the cylinder
II.P.V. either (a) through ports in (l';j). _-/<'/(/.

or (6) ovaonator, such as is used for the condensers

wall opened and covered by the piston,
through valves moved "positively"
by the
that uetnatcs it,
piston or tlio mechanism
or (c) through valves opened and
closed by
in tho oldest
tho oxeeas gas pressure, (c)
to construct,
arrangement and Hie simplest
hut it represents it dopiu'tiire from tho
reduces tho range below
cycle and necessarily
for the connection between
the ratio ;
u^n ceases before the
TJ and tho L.P.V. or II.P.V.
have become equal. It is still
standard piwM.icoinliigh-prossuroeomprc;
IB often
in ovacuators for moderate vacuii (ft)
used in those intended for tho lowest possible
and for working pnonmatio
ono at least of tho valves must be of of steam-engines

pressures, slide-valves are similar to

tubes. Tho
typo (ft),

Throo kinds of piston pump may of a Bteam-engino, but to secure smooth,

between tho
working connection is made
he' considered rather more fully. The Aif/A- is in
shown diagram miitieally L.P.V. and tho II.P.V. when the piston
pressure cowpre-isor, are con-
It would not tho extreme position. Tho pistons
in Fig.I, is nlways composite. to ho
nected in tandem. If tlio pump

be impossible to' obtain a range of pressure it

worked by a reciprocating sl:oain.ougino
would be possible to use the same pmlon-
to avoid
rod fop the driving piston, und thus
and to reduce
rotary motions and bearings
to a minimum but this arrange-
moving parts ;

ment ia seldom adopted; pump and motor

are usually separate.
Tho volumetric offioionoy should
bo about
8fi per cent when ^, = 20 cm. of luorcury or
em. for
moro; about 80 per cent at j)i-===r> j

lower values of p,., 'ICi. will fall rapidly,

of 200 and a atmos.
final pressure of 200 by and ?)], will not be less than 1 om. Tho
over-all efiioiimcy should be ixot loss
than fit)
a simple pump, hut there are several reasons
why tho multi-stage puni]) is preferable, per cent at tho higher pressures.
between IIshows part of a laboratory
Thus, it is possible to" cool the gus "
use. Tho pump is
Intel-coolers 0. cvnc.unlor in Very general
the stages by the Jiy
tho high-pri'Hmiro
tho cranks evenly round the crank- composite with two stages ;

abaft n, more oven torque anil a balanced
member presents no special features and in
not shown it is connected to It. In tho
motion ean bo obtained, Tho construction of ;

low-prossure member shown,

tho piston is
eiioh pump can ho adapted to tho pressures
covered with oil which is ejected at tho end
between which it has to work the thickness

of tlio stroke through tho valve V, carrying

of tho metal can bo increased, as shown in
tho pressure increases j special the aii- with it. At tho sanio time the eraitk
tho figure, as
lubrication can worked by tho piston guide 1C, forces oil
packings and forced
into tho space by means of tho oil-pump H
bo used in tho H.P. cylinders. Some makers

oil as ft lubricant nt high from tho oil flows on to tho top of tho piston
prefer water to
the highest pressures the as it descends. It is claimed that tlio pump
pressures, and at
will attain 10' mm., it tho gas and oil are
substitution seems necessary because oil would
burn explosively. free from vapour. A drying tube with .P B 6
Tho volumetric efficiency of such a punir is necessary in the connection
L to the L.P.V.,
should be wit less thmi 80 per eont ; the gat and another, which can bo filled with ()a(.!t a(
cent ; and tho is desirable in the outlet of tho H.P. cylinder
efficiency also about SO per
ovor-all' efficiency about GO 'per cent. Tho to keep tho oil dry. The pump la very

find convenient clown to letween solid and liquid pistons, thoiigb

of ()!)!mm., but to obtain the highest vauua Hglit from tins Htand point of seicntifif!

is ehai 1

in Jill practical
of which tlio pump is capable needs groat iipld, ]ierfti(;tly

care in its treatment.

In the Torricellian pump the vessel to bo
A piston ])ump for extremely
low profistu rns

]ias been developed by Gaedo

alHo (1) i" 1
ivammlod completely Tilled' with a liquid

not differ greatly from tlmt if density open end in placed beuoath
principle it (loos
iho free surfaces of a liquid, which in usually
;lie same as that filling tho vessel. If, in
his position, nuy part >f
,ho vessel is at a height
tbovo the free liquid mir-
'aco greater than Ii where ,

pf/ (11- T), the surface
if 'the liquid will shik to
tho height A a , and tho
ipper part of tho vesaol
vill contain only tho vapour
of the liquid at tho pressure
corresponding to tho pre-
vailing temperature. As a
iquid for such a pump, j-,.^ 4 _

norciiry is especially suit-

able, both on account of its high density
and small "barometric height," fi ,
and on
iceount, of its low vapour pressure.
This method of evacuation has tho obvious
1'KI. 3.
lisitdvantage that the wliolo
L.P.V. lias to
ie filled with liquid and inverted. A very
shown. In order to free tho oil from water,
ibvioim imidilieatio!) of 1C was described in
and thus to dispense witli a drying agent, into
forced through a special woven
lirineiple by flwedonborg (1722) and put
tin) oil The action
i practical form by downier (18SI>).
material which effects a Hepuration of tho
is clear from ./''iff. * tho ;

two liquids. Gaedc chums thai his jminp

moreury, is alternately raided
without, luiy drying agent will attain a pressure
and lowered and the two-
of 1C" 6 mm.
way cock alternately con-
Aft. Liquid Pinion Pumps nected to the- air nncl lo
is im- tho vessel to lie evacuated.
(15) ToiutinKLT.i'H l?iJMi'{2),(3).It
a perfect fit or a gas-tight
Tho pump is a true liquid
possible to make from
bodies movable relatively piston pump, differing
joint botwocn solid
Cliioricko's pump only in
to each other ; and therefore all truly solid
this nature of tho piston
piston pumps have
some leakage and Homo
and tho valves.
dead space. Leakage can bo wholly prevented
and dead apace very nearly abolished by using (10)TirTPT.miPwMr.
a liquid in plauo of a solid for tho moving
Agreatly improved form
n of tho pump originally duo
portion of the vessel U.
The ratio (p|[ /J>r,)
and to Toplor (4), but realised
ean bo increased by the substitution, anil
though pmetieully by Hagen
higher vacua (or higher compressions Necson in shown in
this result is not so important practically)
fi no stop-cooks are
obtained in a single operation. In fact a ./''iff. ;

It was iitnid in
in this manner in required.
liquid is actually used classical rcHom'ohca
pumps which are usually regarded as of the on low at tho end
Holid piston type. Tn tho pump described
" " of tho nmotocmlh eentury,
last, the oil "covering tho and
piston _
being at that time rivalled only by
tho piston,
passing through the valves is really afl a means of
and the mine purpose is served, in part at Sprenp;el pump (see below)
attaining low pressures.
leant, by tho lubricating liquid of othei is
oE section A. However, the typioa ]}y raising tho reservoir A, mercury
pumps driven up into tho cylinder B, thereby dvivinp
iic|iiid piston pnmjis wore developed historie- out the gas from the cylinder through the
ally from Torriedli's, and not Oiierioko's ho
of evacuation and the dislinetior capillary tuho (',from which it may
method ;

Moreury is
colleiibii'd in tlio mercury trough.
FlKiires In lirnoliotH refer l rpfercniifiH at llio OIK
over into tho
oE lliu urtk'le. prevented from llowiny;

to bo exhausted (V) by the snio.ll glass (17) GAEDIJ ROTARY MKIICUKY 1'irnii 1 .

valvo 1). The general action of the pump can bn Keen

On lowering tho reservoir, Uio mercury from .Fiys. 0, 7, which ai-o vertical sections
flows hack from tho cylimlcr, and when tho paralleland at right angles to thi! front of the
mercury reaches tho lower purl of tho cylinder pumps. The outer casing A contains mercury
gas will cuter 'from tho vessel V througli the to about two-thirds of its height, ami i.s

side-tuba E, ready for expulsion at tho next connected through the pipe B with a rough
stroke, vaninim pump (e.g. an ordinary piston pump)
Tho lower! tiff of tho reservoir muat lie done capable of nmmtaining a pressure of about
very slowly so long ns tho pressure in V is fl!
more tlmn 2-3 mm., otherwise air entering
tho cylinder through the sido-tubo E will

carry mercury violently up tho tube I* and ,

this may easily shatter tho glass-work of tho

When tho vacuum becomes liigli, tho
raising of tlio reservoir must be done very
" "
slowly, otherwise the luiniinor of tho
mercury at tlio lop of its travel in the cylinder 10 mm. An ingenious device shown in ffiff. 8
will break the cylinder head. is used for cutting off the rough pump after
Tho reservoir A should have about twice the preliminary exhaustion. Inside this casing
the capacity of B. Tlio volume of tho cylinder rotates a drum .B, made of porcelain, to tho
B is fixed by the work required from tho pump, side of which is attached a smaller porcelain
but ordinary laboratory models is from fiOO
in drum C. Tho two drums communicate
to COO c.c, Tho capillary tubu is about through tho port D.
800 mm. long and 1 ram. boro. Tho tube I? It will bo soon from this diagram, that if B
should have a boro 12 nun., and fcho lube E is rotated in tho direction of the arrow, the

about 4 mm. This ensures that gas entering portion of B above the mercury will com-
from E to I? will tend to travel in bubbles municate with tbo pipe V, connected to the
upward through tho mercury in P, instead of vessel to bo exhausted, so long as tho purl .1)
carrying the mercury solidly up with it. is not immersed. As soon as .1) is immoi'Hcd,
The P a O B tuba is, of course, inserted to tho communication with V is closed, and when
absorb water - vapour, which presents tlio tho tail end 13 of tho drum rises above
same obstacle to this as to all aerostatic the mercury, tho
pumps. gas will bo passed
Tho lowest pressure which it is possible on to tho rough
to obtain with this pump is determined as in vacuum.
(IB), by the ratio u\,ju\\. u\, is tho volume of In the pump
B, while u\i is tlio volume of tho smallest as actually mami-
bubble which the mercury will carry down the faetiirod two or
capillary C'. This volume is approximately three drums arc

that enclosed bntwnon tlio convex surfaces spaced symmetric-

of two moron ry jrinniseiispH which just touch ;
ally on the same
it decreases with the bore of tlio axle a second
capillary. ;

But there is some gas adhering to the glass drum is indicated

in addition to actual bubbles left behind by by tho dotted lino
morcnry, which makes it useless to decrease in Ififj. 0. Tho speed of tho pump is thereby
tlmt bore beyond a certain limit. Tho lowest increased, but the construction made much
pressure recorded as nltalnod by a To'pler more complicated.
pump is 0'00002B mm, in addition, of course, Tho range of tho Oaodo pump is probably
to the vapour pressure of the somewhat loss than that of tho Toplor, for tho
Tho working of a Toplor by hand is extremely conditions for the expulsion of small bubbles
tedious, for several houra may ho required to from U
are loss favourable. But since p n n
reach tho limit of pressure. Numberless de- owing to tho
is loss, use- of tho
auxiliary pump
vices for rendering
its action automatic have and since its speed is much greater, the least
been proposed electrical contacts worked
pressure practically attainable in at leant as
by the
mercury, or else tho weight of tho low; a pressure of 0-00005 mm. is well within
mercury, are used to- control tho operation. its power. The volumetric speed is not con-
But no description of thorn is necessary stant, as it would bo according to (10) ; it
to-day, for tho problem of the automatic is about 110 om, 8
/sco. at 0>01 mm., and falls
Toplor seems to have been solved fhmlly by of? continuously at tho lowest
(facile fG), who proposed to move tho solid
(18) TUB SPHENOID PUMP. A liquid
vessel mthoi tlmn tho liquid piston.

piston pump, using mercury but working on


a somewhat different principle from those The reservoir E ia normally connected to

just described, was invented by Sproiigel in IT ; an occasional admission of air, in in-dor to
ISfifi (7). It lias over the (.ieisslor (and Inter rcatoro tho mcreury to ."!>, is controlled a by
Toplor) pump the greater advantage that its dimple timing device set once and for 'all.
notion is more nearly continuous and auto- It can also ho controlled by the weight of the
matic, Tlio gas is carried out of tlio L.P.V, mercury in tho reservoir.
by tlio fall of mercury down a capillary tube, The common laboratory mercury still is in
as in tho Toplor pump but it is not forced ; effect a Sprongelpump whereby gases intro-
into that tube it enters tho tube under the
; duced by the mercury are removed.
pressure in tho L.P. V. and is there trapped For high-vacuum work, mercury pumps of
between successive drops of mercury falling these types, with tho possible exception of tho
into tho top of tho tube from a reservoir. In Gaedo puni]), are
another and more convenient arrangement (J^V/. obsolete, and re- |[ H
9) tho drops arc formed in the capillary by placed by those nf
tho entrance of gas from tho L.P.V, through a Class 0. They
side tube. But tho principle is the same ; may, howover, bo
the liquid column in the capillary will break used as auxiliary
up into drops, trapping the gas between thorn, pumps in series
if tho gain in surface tension energy duo to with those of that
tho formation of a liquid-gas surface is greater class when it is
than the loss in hydrostatic energy duo to desired to collect
the accompanying displacement of tho liquid. gases pumped out
Tho precise calculation of the conditions for from a vessel. For
drop formation is complicated ; lint ifc is this purpose
clear that the bore of the capillary must tho Toplor Jj=_
bo below tho limit at which drops could
pump ia most
lio formed in tho tube without completely suitable.
occupying its cross-section. (19) OHHHIOAL
" "
Tho volumetric speed of a single-fall AND OTHEIl PUMPS, I'm. JO.

Rprongol is oxtromoly small, not moro than a The foregoing

few cubic millimetres por second. pumps aro designed to attain low pressurm
It can be increased by connecting But liquid piston pumps aro alno o sorvico
in parallel several capillary tubes for pumping chemically active gases, which
from tho samo reservoir, and
all foil would attack any of tlio metals or other
" "
thus making a multiple -full materials suitable for the construction of
pump. The lowest pressure attain- solid piston pumps. Thus for the compres-
able lixed by tho samo considera-
is sion of chlorine, pumps uro used of which the
tions as in tho Toplor pump, but cylinder and valves are made of load-covered
it may bo increased by small stool,whilo the piston consists of oil or sul-
quantities of air carried into tlio phuric acid. Tho piston
liquid is Hot in
vacuum by tho stream o mercury motion either by compressed air or by a solid
from the reservoir exposed to tho piston working in one limb of n IJ-tilbe, tho
atmosphere. Many devices liavo other limb of which is tho chlorine pump.
been suggested for avoiding this Laboratory pumps essentially similar In tlus
defect. (See (,')).) Toplor, but using oil or glycerine or sulpburii!
Tlio Rprongol pump ean be made acid in place of moroury, have also boon used
completely automatic, if it ia for some purposes. The lower densities of
arranged that tho mercury which theso liquids enable tho pumps ti> be umdo n(
has fallen down tho capillary IB glaast with a volume much greater than that
restored periodically to tho reser- sot by tho moolmnical strength of tlio glass if
voir. Such an automatic form was at one mercury were used or bho gun to be pumped

time in universal use for tho exhaustion of may bo ono which attaolm mercury.
electriclamps (8). It is shown in Fig. 10, Again, air compressors for high prefmuroH
and the method of operation is obvious from have been constructed in which water is used
that figure. as tho piston in order that the cooling of tlio
Tho outlet an auxiliary
is connected to gas may be more efficient. I" some of tht'so
pump maintaining a pressure of n. few cm. mctal chains hanging into tlio water from the*
Tho capillary fall tubes A, into which mercury top of tlio piston have l>eon used to faoilitato
flows from 1) through the jets 0, aro therefore tht! transfoTonaa nf boat between tlio gas and tho

relatively short (about 20 cm.). Tho bent liquid.

tube '\\ enables the end of tho exhaustion Ac.. Flexible dan lantern
to be seen by tho disappearance of gnu-bubbles (20) BELLOWS. In theso pumps tho vessel
from tho meroury. U has flexible walls and Ha volume ia varied by

of the Tho board B fulls by its own weight and_ the

changing its shape, Tho advantages Air is
due to air outers through the fiat valve V,.
typo arc high mechanical efficiency in the feeder by a hand lover,
tho ftlHonco of friction, simplicity of compressed
and reliability. and the air is driven through tho valvcn V a
tion, and consequent cheapness the
into the reservoir bellows I), from, which
On the other hand, they have a small range Tho pressure on l>
tho ratio MH/MC, air passes to the organ.
of prcasure, partly because W on the top board.
is comparatively groat, partly
because tho depends on tho weight
a uniform pressure in .1) whether
makes it impossible to To ensure
flexibility of iho walls
from is expanded or contracted a double set of ribs
nan them at pressures differing greatly outwards
made for hand- is used, the upper set R, folding
atmospheric. They arc usually tlio f ratlin
or lew mechanical drives. and lower set ll a foldin" inwards ;

working very power

tho H. between the two sets of ribs is connected
Tlio earliest 'examples of tho type are halves
lunga of air - breathing animals ;
tho later by a mechanism shown so that both
of tho of the reservoir bellows expand equally.
originalmodel for this {22). At the opposite extreme of the typo
tho sqiieo'/cd rubber tubo pump (./''if/. IB) 0").
pui'poao arc insignificant.

The type is almost This pump consists of a rubber tube A,A|

B and
equally familiar
in bel- , 5) wrapped inside a hollow cylinder

lows of all lands, for squeezed by two or

more rollers 0, and
blowing up a lire, for
vacuum cleaning, for so that tho way
piano -players and squeak- through tho tube is
nt tho
ing toys ; in the fountain- stopped
pen filler, nowpartially squeezed portion.
a Tho rollers O and As
replaced by piston l

Oj roll round
in the bulb for the
Fiu. 11. pump ;

scent and other sprays. inside of tho cylinder

driven by the shaft
But it is also used for less commonplace
1), and gas (or liquid)
is is transferred from
The action of Iho blacksmith's bellow^
shown diagram matically in Fig. U. The A! to AS as the shaft
end plates A and C arc fixed, while B is given revolves.

EL reciprocating motion.
Some bellows of tins Tho squeezed por-
feet in tions act us pistons,
kind for smiths' hearths are several " "
of leather. and these pistons
diameter, with the flexible sides
air is usually pumped into a
reservoir arc formed at AI
The A2 where they
bellows in which a pressure of about
in. of and travel along the tube to

water ia maintained by a weight. disappear.

on a different scale, The action of the pump is somewhat similar
Fig. 11 also represents,
tho rotary pump of /% 10, with tho
a useful laboratory imnip. Tlio bellows arc to
" "
hero made front tho inner tube of a motor tyro, important di'ffercnw that tho piston
tho corrugations being obtained by largo metal the rotary pump requires to be carried across
and small rings outside from A, 'to A and is thus liable tn causo
rings inside the tubo

Tho cylindrical discs A, IS, "of air or liquid.

leakage Tho tube pump has
placed alternately.
C are of aluminium, and U is driven by a crank no dead-space and is only limited in range by

Irom a small motor, tho strength and tightness of tho rubber tnhn
The most of this to resist the pressure difference hut its speed
elaborate pump ;

nlaaa is the organ bellows, shown on Fig. 12 ; is small and mcdianiimlly it is inefficient.
the bellows com- Tho tube pump 5s particularly mutable fm-
pressor, called transferring gasca or liquids without mm-
" "
tho feeder lamination, as tho plain tubo can bo eamly
hollows, ia at cleaned and no other portion of tho pump
F. The upper can come into contact with the fluid being
board A
of the pumped.
feeder is fixed An ovon simpler pump of tho name type
and the lower can clearly bo made out of a plain piece of
board Bislnngcd rubber tubing pressed with tho fingers,
to A by n. leather
Tho Ad. liotari/ Pumps
wedge-shaped (23) BLOWERS.-Tn this class the variation
volume between A and 11 ia enclosed by the of the volume of U
is effected by tho rotary

wooclou ribs R ; tho riba are hinged to each motion of solid bodies constituting part of itn

other and to A and Hwith leather and cloth walls. The pumps arc usually driven by

(-1 are intended for continuous action ;

icnl. The over-nil mechanic-ill efiiciciioy is

Joto therefore- with, solid piston pumps abimfc 7 per cent.

tn with other types of Class A. Over Fty. It) shows another whero the ly]io
pumps they have the advantages imjidler vunes V are ffialc*iHHl (it
one end to &
disc whicsh Cannes them to rotate around the
-nical efficiency ;
in incchanieal

and in cost, both fixed core. The vanes nltrr the delivery
niicl for upkeep ; in e compactness ;
sti-olto eomo intu the openings in llio rotating
noas of air current when used as body A (called the id lor), which is canaccl to
rotate at the same speed as tho vanca b
They have disadvantage in a smaller
leakage, and on the outside. On rotat-
liiTsauro, in greater
iti noisiness. But pumps of this ing further the vanes como
also with those of Olass 1! into tho suetion chamber*
V M"' blowei-s typo lif/ whence they start again on
1ms the advantage in mechanical the compression strolio.
ami simplicity. Ifor eompressnrs Tho pump is more com-
little to choose between Ity and plicated. than the Hoot,
but several advantages arc
though tho latter has tho greater surfaces
th types would bo replaced by AH claimed. Tluia . If).

;ov range were required. evacu-

l)'or ean bo used to separate- tho
is useless, while Ad provides the two olinmbers whero lines only are possible
in tho Hoot blower ait- is compressed by
machines of modern practice for all ;

there is no contact
between 10 mm. and -01 mm. one rotating part only ;

Kief examples of the class can be between parts moving with different velocity,
mid tlniH there is IOHH friction the mechanical
into two groups, one (I) developed ;

construction and cheap tho pulsa-

from the liocit Blower, the is itii])le ;

other (2) frnm the IJcaln tions of gas arc redueed.

In another typo fhe rotating parts are
Blower. Tho development
npiral vaneff, which give
a more even, delivery
lias been so gradual that it
associate any of {jus uiul makf] ICHH noises the mechanism
is difficult to

of tho pumps, or even the in not .easily nlinwu

in a diagram. Many
deviees liave been Rfimo vary-
two archetypes, with the (itiicr ui!ii])te(l,

ing widely lit detnil

from thosp mentioned,
.. name of any inventor. The

but all based on the samo principle descrip-

groups are" usually distill-

tions o them are to bo found in maliera'

:yy tho nature of the abutment,"

iiio or siu-faeo dividing tho 1T.P.V. ciitiilogncs.

These! puni]iH, of
D 3'j.P.V. group (1) is then tslnirae.
(25) Fix Kn AntiTMHNT.
movablo abutment, (2) by it
a which tho licalo blower in an early example-,
are used extensively as compressors, aa blowers,
ittmcnt; for though in (2) the bodies
tho abutment move, tho lino or wir- and as ovaomttors. They are mcA in Rns-
es h is tho abutment is a.t rest relatively
workH for

pumping the
LE ABUTMENT. TllCSO pumps gas: to the p
holders, and -B
fly used for moving largo quantities
tho factory or
against a small pressure difference

to fi-10 feet water pressure). laboratory for at-

vly example typo was exhibited
of this taining pressures
J?aris Exhibition hy Elilm Hoot in down to -001 mm.
It consists of two two-tootlicti Tho general prin-
ff, 14),
A, :ii, whioli are made to revolve at ciple employed
o rate in opposite directions by moans shown in Fitj. 10.
tmtsklo the box or pump hotly the ;
Tho cylinder I) rotates alxnit anaxiHl),Bo lhat
it'tonohos tho oontainhiK cylinder C ut tb
iciludcd botwcon tho wheels ami tho
is U, and its volume varies with tlm abutment 11. A
Hint in 1> curries tlw llnlc
" cuds of which ore held
of tlio wheols. If they potato as or sorapoi-H," the outer
C. Tho ptalcs
11 the figure air would ho sucked in at against tho containing cylinder
divido th aiiacn Iwtweon and I) into two
eliverod through '!>.
the volumes of thcao two
-ovent louUage tho wings .are muehined parts ; as 1) rotatPH
often in n inannor readily ncen from tho
.trntely an jinssilile, and tiro Itarta vary
in which P is tho auction imd Q tlio
with wood or other packing material. fij^uro,
small press- oomprMsion inlot.
ImnoU'ic oflleieiuiy against
inolies of water) may bis as Tho friction of tlio snrapors on thu cylinder
ny 10
involves coiiHidorabb lows imd \ITIU-, and many
Otl cent vi'hile nt higher ])ress-
t ]ior h

ft-ct of water) it will drop to 80 pot- alternative arrangcnienta have been devised

bo avoid ife. In one, contact between tho that the fluid is incompressible,
or that Un'
scmpers ami tho cylinder in preserved by a change in pressure is infinitely small (2) Lhivl^ ;

emiular guide with its cenlro (joiiusident with there is no loss or gain of energy to or from
that of 0. In another thorn is a single solid other sources, c.y, friction of the moving gas.
scraper which slides frcoly in tlio slot in 1) ;
.By tho pressure " must bo understood tlio
fcho section of tho oylindor 1) is not circular, force per unit area on a surface at rest relative
but such tliat tlio two ends of tho semper to the gun in a frictionless fluid it is eqmil

" "
aro in contact with tlio walla whatever the at any point to tho static pressure on ii>
position of I).another tho sonipors
In tiro surface parallel to tho flow at that point, bill'
on nil " '*
pushed out by springs, but they boar moving relatively to tho gas ; tho dynamic
idly rotating cylinder (" litting closely within pressure, or that on a surface perpendicular
O nnd pierced with holes ; tho clearance to tho (low and at rest relative to tho pump, in
between C' and 0' is inado no small that tho p + lpv , n,] id, when (24) is true, is constant
leakage between the two is inappreciable ; along tha whole stream.
tho friction is thus reduced to that of 0' on
its bearings. In any lie. Injectors and Ejectors Qaaeaua Stwam
pump of this typo, tho {General Reference (11))
number of scrapers may
bo increased. Fig. 17 Those are wholly analogous to liquid
shows a typo witli throe jot "Hydraulics "). A gas or vapou r
pumps (see
l< "
or nioro scrapers, hinged (called fluid to distinguish it from the*
at tlio central axis of M gas to bo pumped) is forced through a tubes
J tho box IS, mid sliding
N from a reservoir 11, at pressure yt\\ into a. t

FIG, 17. in cylindrical stuffing larger tube communicating with the atnid-
boxes fastened to tho sphcro (aao Fig, 18). If tho How satisliml

rotating drum D j this drum touches tho assumption (2) above, tho stream would nub
bo brought to rest in tho atmosphere unlcnH
easing in tho fixed abutment E.
In pumps of this typo designed as ovaouators p tt wore equal to II, and if jin = ll there woulil
oil is always introduced into tlio cylinder to bo no stream. But owing to viscosity and
prevent leakage and to fill up the dead spueo, friction, ?)]j may bo greater than II, so that u
Tho scrapers aro usually arranged aa in Fig. Ifi. high volooity i>x is obtained in N, and yet
In order to prevent hammering by the oil may tic zero when tho atmosphere in reached.
when tho vacuum is high, a valve ia fitted In these conditions, tho clifToronco 1 1 - p, whero
which limits tho quantity of oil flowing from p is tho pressure at N, is not HO great as Jpl'.v 3
the compression side it also helps to sopamto
; given by (24) ;

the air from tho oil. Such ovaouators aro but it is still
often run in tandem, or one of thorn is used in finite nnd of tho

series with an auxiliary pump of some other same sign, j p is

kind. If pnh'i to 10 mm., p { nmy be reduced leas than. II.

to -0001 mm. If ga-R in the

Semper vacuum pumpa am largely used space surround-
aa auxiliaries to high- vacuum pumps (C) ; ing N ia given
they aro also tho cliiof typo employed in tho across to tho
preliminary evacuation of electric incandescent stream through
" " tlm gap between
lamps, which aro subsequently clcaned-up
by tho discharge (D). N and (I, it will flow into tho stream of fliiM
and bo carried away by tho stream, BO lonj^
]J. AKHOOYNAMIO PUMPS as its pressure is greater than p, Tho iluitl

(2(i). In aerodynamical pumps, tlio proas- streaming from N to will suck gas throng] i

urofl aro functions of S and of tho the pipe and

will aet as an ovacuator or
velocity " " " "
of the gas in different parts of tho appaiutiiH. In an
ejector. injector or compressor,
Tho fundamental connection between tlio Q communicates with tho atmosphere, and thi*
pressure and velocity of any fluid is given by space with which communicates, and in
tho familiar hydrodynamicat which tho stream comes to rest, is at a proHBiim
greater than II, but, of course, still much ICHH
%H-^ a
)^0, . . .
(24) than pit- ]) is equal to p ff in an ojoctor and
or Pi- P = &(* -Vi*}. - -
(2fi) to p,, in but if there in u
compressor ;

It deduced on the assumptions (1) that tho

is 1
OonfiiHlan IH somntlmca lulrndimert l>y n fnlliiro
energy required to change the
pressure of u to observe Unit tho chief part of t!m nrcflsnro of u
volume Y of the gas from ^1 to p a is (p - Kan IH iiiMopaml)ln from ItH comim-aslblllty It in
l p^V, not ilno, liko that o( a liquid, to its welffhl. Thu

(2) that the energy of any such change of lircssure of tin Incompressiblo HIIS Is a nicnnliiRlmH
pressure which occurs is equal to the change (toiiceiitioji. Tho fliiplliifitlon of the theory to RaHC'
in tho Idiiotic energy of tho gas.
Is only becnuap, for small ohaiiRca of iiressuro
(1) implies lit constant temperature, pdV = - Vilp.

o on tin iml
stream of gas through the pump, g (28). fn practice tho fluid used ift generally
pi,or PH will not lie equal to j>, (1) because of strain or if((iii]ii'(.isscd air. Ejectors using tliRHo
the drop duo to How of the connecting tubes fluids are n.ied for vaeinun brakes, vacuum
(J and 0, (2) because tho mixing of Urn gas cleaners, and grain conveyers. Their great
with tlio fluid affects greatly tlio velocity mid advantage) is, of course, their simplicity and
pressure of the latter. freedom from maintenance charges, fig, 18
In calculating the performance of a pump, shows a pump used for railway vacuum brake
allowance has to bo mado for departures from operation. It will rnduco ?>j, to about 15 orn,

(20) owing to friction and viscosity. The of mercury. A more elaborate pump is
allowance ia usually made by introducing shown diagrainmatieally in Fig. 19. Hero a
on the right hand of (25) an empirical factor
f, less than 1, and writing

where v lt v 2 are tlio mean velocities over a

cross- section of tlio stream. The principle of
the calculationis thon simple. There are throe
equations (20) for tho tlirco tubes N",' 0, (J ;

and there is the equation of conservation of

mass when tho streams meet. (Tho momen-
tum is not conserved, for thoro in a reaction
on tho tubes.) Theso four equations suflieo
to determine tho four unknowns, viz. p, %,
uoi f<ji in terms of p\i, pj, t p\\, tlio densities
p e and pf of tho gas and fluid, the throe
empirical constants for tlio three tubes, fa,
i'tit fa, and I (I
N, fc\)t -l^q.
t'bo cross -sections of
the three tubes at their openings. Tho algebra common of steam works two pumps
need not be set out, for the numerical re- in series. Tho lirat consists of the plain
sults depend wholly on the values attributed noM/.lo A,tho stiiiond in tho annular gap .11,
to the empirical constants it in given in (11). ;
from which the gas is carried into tho sur-
Here it will suffice to state some of tho most rounding nnnnlar space ('. It is claimed that
important qualitative conclusion", which are this pump will attain n pressure of 3 urn. ol
confirmed by experiment, These wore first Kemark Hhould lie iniido that tho
stated by Uoiinoi (12). 1

the simple theory to sueh

a]i]ilication of pumps
Tho variables considered are p\{, pi., pn, %, in

extremely precarious, for the assumption

I^N. l 'o,

I'V pa , />/.
iwul 8, the volumetric that tho change of proHsniro f the gas is

speed, which is equal to Wqli'tj. In each infinitesimal is clearly [also.

statement tho variables not mentioned aro UO showa a blower usort for
Fiff. moving
supposed constant. largo quantities of air in ventilation under a
(1) P-Pi, is proportional to pn and %*, pressure of a few
HO long as ]}[i in great compared with pu or p Sl . inohea of wator. Air
(2) proportional to \^pu, to UH, am! to
8 is or steam ia used as
1>\\ subject to the same condition.
-'/>L Iluid tho ooncontric

(3) depends only 011 tho ratio.i of

PnPi, cones ore designed to
tho i'"s, and not on their absolute values. make tho velocity of via. no.
(4) S is proportional to tho !?", if their tliogas Hourly parallel
ratios are constant. to that of the fluid, so thnt the dii'ectiim of How
(fi) Given one of these ratios, there in an of the latter is not disturbed by irregularities

optimum value for the other two, giving in tho flow of tho former. A somewhat similar
maximum S, but the same maximum S oan adopted m the smoke-lmx of

arrangement in

bo obtained with different values of tho ratios. a locomotive, whore the exhaust steam is
(d) S is independent of p a and pfl so long as miulo to create draught through the linilcr

VH is constant. duos.
(7) S may bo considerably greater than the (20). Tho most modern development of
volume of fluid insuring nor second from N, tliotypo is tho mercury vapour jot pnmp used
" "
two or three times as great.
e,g, in conjunction with condensation high-
Owing to the niroiinistaiiecs in which the vacuum pumps ((/.!'.). Indeed, as will bo
pumps are used, the efficiency is seldom seen, the lino between vapour jot pumps and
important. But the mechanical ellicic.ncy condensation pumps cannot bo drawn sharply ;

reckoned on tlio basis of the work done in roughly it may bo .sot at tho pressure where
dr'.ving tho fluid stream appears fmldom, if the mean free path of the vapour molecules
o ir or, to exceed 2/i nor cent. becomes comparable with tho diincnniniiH ut the

tulies ; but there is no justification for over- calculations confirmed by experiment scorn
looking biio distinction entirely. Tho isonstruo- to have been based on any view, and few data
tion of these vapour jot pumps is essentially of performance or of. its variation with the
similar to yiij. 1.8, but the apparatus is made construction of tho pump seem available.
of glass vapour
; the stream ia producer! (vJl). Tho filter pump of /'V.?. 2U fed with
by boiling mercury, and confirming arrange- water at a head of BG ft, or more will reduce
ments aro provided for returning tho vapour vapour pressure of tho water. Jinfc
pi, ii> tlio
to tho [toiler, A practical form designed by ti<> measurements of 8 under varying conditions

Volmor (13) will reduce the pressure from have been found, 1 1 is recorded that the,
= 20 !"' mercury to jJi, = '001 mm.
() f pump is more efficient
placed at tlio top

iiut it is doubtful whether they will replace of a building HO high that tho exit tube can
generally the rotary aerostatic oil-pump for bo made as long as the water
producing the axixiliary vacuum of high- barometer.
vacuum pumps, Fragility is their great fault. A variant on the usual design
is shown in Fig. 2 IB, which is
B/. Injectors and
Liquid Slretun Kjectors similar in construction to the
(30). This typo of pump, of whioh tho Vontiu'i meter. But sinco fclio

laboratory filter pump (ffitj. 2Li) is & common notion is improved by a linffla

example, is often regarded ns a mere modifica- at 6 which breaks up the stream, 2111.
tion of .Be, gins or vapour being replaced by a itia probable that, as suggested,
liquid as fluid. Iiut tile difference is really the formation of drops in an important part of
great oi .
tho theory of lift is applied to
1C tlio process. A
imn-roturn valve, as shown at
pumps with liquid as fluid, then, even iE all V, is tiscful with cither of thcHO types to
phuiMhlo corrections arc made, tho prevent the How of water into the apparatus
calculated pert win once is far lass if fcho head becomes insnfllek-nt,
{&.g.10 times) than the actual. Tlio JUwiouatora of these tyjics nro applied
error arises in assuming that the outside tlio laboratory to vacuum cleaning
gas and fluid are miseiWo. The anil to grain conveyors, Compressors work-
flow of tho gaa into tlio fluid is not ing on tho same principle, but with a difl'oront
determined simply by the pressure- construction, havo also important (lommoroial
difference, and relative motion of lines ; they aro known as brompes," In a
tho fluid and .gas ia possible, ovon very simple form (fig. 22), used for blowing
niter they aro mixed in the exit

blacksmiths fires, a stream

tube O. of water flowing tlowji a
Ik sccina preferable to look ufc
2lA. pipo with 11 low holes in it
their action from a different point drags with it ir from tho
of vioic, TITO processes are involved : first, atmosphere, wliiali is fldlj-
tlio entiuiglmg of thogtts
Iiythe liquid stream; eequoniiy separated from
second, tho conveyance of the entangled gas tho water in a closed vessel.
from the LJ'.V. to the H.P.V. During the A more elaborate form has
second jii-ncesa tho gaa will move relatively to hccm developed in America
thfi liquid nearly as if the lirjuid wove at rest for mipplying (iiinijtrcased
i-cliiUvo to tho vfiMfi.
prossuro The (lilVni-nnco of air to mines whero a great
ultimately obtainalile is limited only by the hojwl uf wnicr is availftl>lo,
condition that the velocity of thn Jiqiiif) ontcj-- Tlio pressiti'o o)>taii)blo
ing tlioI/.P.y. Jy N is Buffioicnt to orry it

i o eonsidorablo frnetioa
ift tllpoiigh O flgatust tlio prcsstiro 'ji\i-jij., nt tliat cm-ri-'spomling to tho head of tlio
and is also greater tlnui the velonity with water.
which tho buliblcs of entangled gfia travel
Ihrough tlio liquid in tho opposite direction.
It is tlin second process which determines! (General Koforrmco (II) and (14))
fch,o greatest possible value of .pjj -prf, (32). Centrifugal air-pumps are analogous
On the utlior hand, the apoecl of tlio pump toeonLrifi'gallin_nid.pnmps(seo "Hydraulics"),
is detcrmincfl by tho first process. Its nature "
They nrn gcnoi'nlly callod fans," fttid urn
5s obsoiiro probftWy tlio )ir[tji(l stream carries
; used as lttnn or blowers according to the
along a layer of gas on its surface, in virtue of definition of tho introduction. Tlio principles
friction mid viscosity (of. (39)), aa would a is shown
in Fig. 23, which illustrates the

solid rod travelling with tho stums velocity. simplest typo. Tho gna ontormg the eiroiilar
When fcho liquid breaks into drops in virtue central aportnvo in tho housing is whirled
of the inherent instability of liquid jots, this round by the rotating vanes, acquires velocity*
gas beenmoB entangled between tho drops. and inso8 through Ii.
On this view tho jwrfornmneu of a pump Suppose that tho conditions neco88a,yy /or
of this Typo appoiws quite incalculable. No (24) aro fulfilled, and that tJio gA9 loaves t),o

HM volooity (0),
HO that if w is tho power exerted at tho
tipsVTho vanes with a = &(pn~l'>j )lw. Some-
J, pump shaft, 'K, uw
of the oirculai \,. l

uniform over tho whole surface times tho useful work is reckoned hy ())),
the vanes, llw area
oylindor surrounding corresponding efficiency is called tho
surface is Zllrf^i,
whore r m tl w tho
of this total efficiency, hut it is seldom important,
l)readth poi-
radius of the vanes and d their
previously explained. ~K, uce^. IK a
as true
the diagram. Then
pendicular to efficiency and can never he greater than 1.

s=li liv
. . .
(27) Tho volumetric efficiency K^,, which

and i-a is tho is taken hy some writers to bo S/ty ftn(l by

If Fa is tho area of the opening, others to bo S/Z-n-rdv u The latter quantity

of exit uniform over tho opening,

would bo unity if J\ were equal to v tho ;

G "17 ?) . - \*-"/ former seems to have no general significance,

hut, being a no-dimensional magnitude, is
its velocity from the
If the gas in acquiring convenient for comparing similar designs.
ta tlion
vanes preserves its original pressure p E vn is often greater than 1 if tho first

from (2fi)
expression is chosen, sometimes if the latter
jSa (29) is chosen.
p -3, 4 =ip(V-"a')=i' (ir*"F")'
It is apparent from (30) that tho pressure
is not true and and tho efficiencies will vary with S, if S is
But, as in Section He, (24)
allowance has to he made for
losses of energy controlled hy changes in the area V 2 hy m
duo to friction and to sudden other changes in tho resistance to tho flow
changes in the direction of tho gas. Fig. 24 shows typical curves
tho gas stream. Further, the relating tho brake H,P. w,
tho pressure clif-
velocities are not uniform Over fcroneo p^-pi,, and tho mechanical efficiency
the surfaces l!\ and liV Again, to the volumetric speed S, the velocity of tho
it is convenient to express vanes heing constant, It will bo observed
PH-PL '""I & Ul terms of tho that HUM efficiency is oro for S~0 and for
high values of R and has a maximum
of tho vanes which can for some
1>o measured directly ; it is usual to represent intermediate value. If mechanical efficiency
required tho pump must be designed
of the for its
this velocity hy v a tho linear velocity

ho assumed that v v and v are If the velocity of tho pump is

tips. It may special work.
proportional' n
Tho losses may bo then
to i' . varied over a moderate range, S varies as
terms proportional
represented by one or more
to v\ or to S
or to Sv and tho gonoral equa-
| ,

tion for tho performance of tho pump written

_ (U ,
S+ .

v Ra. (30)
,, ,
o 2.|.p ? Q

The constants a, 0, 7 are usually regarded

at which tho
as depending on the angles
stream of gas strikes tho vanes and tho
on tho
housing they certainly depend

of tho
geometrical quantities characteristic
pump. Some progress towards calculating
thorn directly from those magnitudes can
made, but some purely empirical constants is
Pn-pr, its v and tho power "expended "
are always necessary. In designing the forms Vj,

For extreme varieties tho constants

of the vanes and of tho housing such calcula- ".

tions arc a useful guide here reference can ;

of (HO) change.
differ in sr/.o,
made to discussions in (11) and (Id). (;!4). Simple centrifugal fans
only bo of the vanos, and in
ho noted that in (80), is always, p in 'the number and shape
It may the shape of the housing surrounding
usually, positive,
while 7 is negative.
Three kinds of efficiency are recognised for This is often divideil into a "diffuser,"
sides immediately out-
centrifugal pumps : portion with parallel
or portion
'.['lie inanomotrio efficiency T!,,,,,,,.
sido tfie vanes and a "volute,"
g (;);}}__(!.)
of circular section, outside the diffuser.
is taken as (j) H -?)i,)/p
,?>i, (usually If) and p n
increases towards
tho cross-section of tho volute
being tho static pressures of inflowing
use of tho
and outflowing gas. If were equal to , tho outlet in order to make some

()) Somo fans have

tho maximum value of IC|, I!H1 would bo -J-,
kinotie energy (see
on both sides of tho fan, some

but since w, may bo either greater or ICHS than inlot openings

have any value on one. JUit in their performance,
I'-TILUI, might theoretically
; only
it is soldom if over greater than I .
they all have common characteristics they ;

The mechanical oflii'.ionoy ]<!,,,[,., The are' all used as low -pressure blowers, the
obtained being about 12 m.
useful work is generally taken to ho given by maximum pressure

and to the cube of the velocity, (:U>)


not exceed 1-03, power

water palpi, does
true, at least approximately, and wimiliii'ly
of ;

of is . In still
assumption (1) (28) stiilo
J^tihe, may defined efficiencies might be employed to
large sizes their mechanical efficiency
it is more nearly the performance.
reach 80 per cent, but usually The difference between airscrews and
between that
70 iier cent and intermediate in tho connection bctw^n
Root's Blowers. Over centrifugals lies
of piston pumps and of constants of these equations and fch
have the advantage the
cither of these types they Much more is known
hot geometrical magnitudes.
of simplicity and of being proof against on
arc suitably of this connection for airscrews, perhaps
and diisty gases if the bearings
.account of their im-
Composite, or multi-stage, centrifugal
fans portance for
other pur- "'-
" Mt ' ^ '- '
Ss" *

common. The fans of successive poses; for this know-

are also
on the same shaft the discharge ledge reference may
stages run
" Aero-
of one fan is led by a bo mado to
from tho circumference
tube to tho central intake
of tho next. Since dynamics"; since tfio. 20.
airscrew pumps are
- Pi is proportional to p by (25) and p is tbLIOULUll |_Fiim_|-
account >i

in successive not very important, no further

to p,. or p a ,vn have bore.
proportional K 1S 1G their theory need bo given
or venM-
stages Sp'p,
is the, number ,
(36). Airsorows are largely used for
range of the simple pump up the air in a
room Mi-
pump totion, either stirring
of stages The range of the composite u
of the extracting it into the atmosphere through
the" product of the ratio-ranges

hole in the wall. A wi-wii-ov is visry
individual stages. Such multi-stage fans with plain
since th
* = 1-1 when inefficient for tlio second purpuw,
ten or more stages, each giving the centre iul
4000 difference of velocity between
driven a turbine or electric motor at
by a cLrmilii-
used to deliver
are circumference of tho screw
tion within the fan itself,
as shown in Intf. ^
is greater wlmn,
ail to blast
The loss' due to this circulation

as in Fiq. 26, it is the H.l'.V. that

is partially
furnaces at a
closed than when it is the L.F.V.
To nlm-
pressure of 2-J-
is ofbnii
atmoa. They the loss the centre of the airscrew
with a disc toprevent the rotum
are also used covered
for "
Kiipor- flow; the volumetric spued for a
charging" diameter and velocity is tlmrobj
petrol motors
but tho mechanical efficiency in increased.
It is to secure that all the en
on aeroplanes. impossible
shall produce axial f
In Fiij. 2fl a given to the gas ow
find radial
is shown details of some inefficient tangential
3-stege fan of this kind Jn

same tune.
and the is always produced at tlio
bearings, rings to prevent leakage, tlio
outer casing have been tho Ratoau scrow fan, shown in Fig. 27,
water-jacket of the " flow
im- tangential and radial
omitted. A, B, C are the revolving
tlio parts drawn solid are
fixed. ^sgreatly reduced by causing
pellers," while
the gas 'bo strike the blades
Bateau constructed a simple fan, running
at 20,000 r.p.m., which gave
= 1'6; experi- (B) with a velocity opposite
to that of their rotation.
ments on extremely high speeds have also
This velocity is imposed
been made by Parsons and others. But
such fans seem to have no practical
on the inflowing- gas by the
simple The centre
fixed vanes V.
advantage over the composite type. covered by
of the blades is

the fixed disc T>. ft shows

B/i. Airscrews
u transverse sec ion through

" "
(General References (11)
and (14)) the fan, b a cylindrical
the velocity of tho gas section made by a cylinder
g (35). In type %,
coaxial with tho fan cutting
produced by a" rotating solid
is perpendicular ,

to the axis of rotation ; if tho velocity is

15 and V
and developed into
tho fan may a plane. In 6 the motion of
mainly parallel to that axis,
bo called a
" airscrew. the blades B is upwards. The Katean wvrow
propeller," or, better,
In all that concerns tho general relations fan resembles in its performance a Himplo
between tho velocity of the solid and tho centrifugal. "
the airscrews Some fans, described as of mixed flow,"
velocity or pressure of gas,
are indistinguishable from centrifugal pumps. aro intermediate between centrifugal am)
Tims tho pressure produced by an airscrew airscrews, tho flow of gas boing partly nulial
or tangential and partly axial. lint they <ln
of its velocity,
is proportional to the square
the volumetric speed to tho velocity, and tho not differ sufficiently in principle from the

many fcypos of pure airserows and centrifugals,

wlticli are also described in makers' C/. Friction Pitnijis
to warrant special notice. .For small powers (30) GAKOB MOLECULAR Pump. The
there Booms little to choose between these action depends upon tho forces between a
classes of fans ; for larger powers the contri- gas and a solid (or liquid) aiirfaeo moving
ftigal is more suitable ; it is also more suitable relatively to it, At ordinary pressures these
for the individual members of a composite forces are determined by tho
viscosity of the
pump. gas, and the influence of the solid boundary
enters into tlio calculation of tho flow only
Ri. Thermal Pumps
through the assumption that v n tlio velocity ,

(37). The principle of those is sufficiently of the gas at that boundary, is zero and that
dismissed under "Convection." The chimney there is no "slip." But at sufficiently low
of the open fire which ventilates a room and
pressures Kuiiclt and Warburg (15), continued
tlio gas jet in the fluo oftlio chemical fume
" " by many later observers, showed that the
cupboard are familiar examples of blowers measured (low agreed with that predicted
of this typo. The draught produced by a liydrodynamically only if it was assumed Unit
flame in a fluo has also boon used to work there is Home slip, that v v the
velocity of tlio

small wind channels for aeronautical investiga-

gas at the boundary and parallel to it, is
tion. and generally, if only very small powers
finite, and that tho force exerted on tho gas
are concerned and efficiency is unimportant, tlio
by boundary is cv u t is called the friction

chimneys and small fans may bo regarded as coefficient and e/i; the coefficient of slip.
mutually interchangeable. From the molecular standpoint the matter
appears somewhat differently. Tho condition
C. HlOH-VAOUUM PUMl'8 i>= menus that the velocities of tlio mole-
cules leaving tlio boundary are symmetrical
(General Refenmocs (21), (23)) on either aide of tlici normal!. The aj Clearance
(\}S). During this last few years pumps of slip atlow pressure's dues not mean thai this
1m vo been invented which will attain pressures condition is no longer fulfilled. l'r if the
lower than those that can bo reached
doliiiifcoly pressure is greater on one side of the normal
with any of the pumps described so far. They than on the other, more molecules will arrive
" "
dopcmd upon molecular processes, that is at Iho boundary from the first sido if the ;

to say, processes explicable by molecular molecules leave tlic boundary equally dis-
theory and not by hydrostatic or hydro- tributed on both sides, then there will be on
dynamical theories, which regard a gas as iv tlio whole* a flow of gan from the first side to
continuous medium. These processes become tho second, so long as the distance travelled
important only when the pressure of tlio gas ia by tho molecules leaving tho boundary before
boknv some dolinito limit, which is usually far they oollido ia finite. On bho other hand, the
liiilow that of the atmosphere. Tlio puinps flow will bo lesHs than it would bo if tha mole-
must therefore bn run in sorios with an culcs tho boundary with their velocity

auxiliary pump .whioh roduoos and maintains parallel to

it unchanged.
Accordingly the
tUo pressure below tlio limit at which tho condition ~0 is not inconsistent nt low
notion of the molecular pump begins this ; prosHiires with the hydrodynamical assumption
pressure is of the order of 0- 1 mm. As of a finite slip coefficient.
auxiliary pumps, those of typo (Ad) are now Knudsen (10) hus calculated tho friction
usually employed. Further, since the vapour coefficient from such a molecular theory. Tic
pressure of water in much above the limiting assumes that whatever tho direction of the
pressure, a drying agent must bo used in incident molecules, the mini tier with any
conjunction with the auxiliary pump on ;
velocity leaving tho boundary within a cone
the other hand, a molecular pump does not of solid angle ilia malting an angle with the II

distinguish between vapours and permanent normal is proportional to eon Cdu and tliat Iho
gases, and no device is needed to remove distribution of velocities in Max well iail. Ho
from the low- pressure side of the pump any concludes that e--=e u .ji, where
vapours except those whioh arise from tho
tic Lion of tho
pump itself, i

Two molecular processes have been employed ;

for mush pumps, both originally suggested by and p is tho density of the gas at unit pressure
Gaodo. Kinc.n the lirst typo was tlio only and tho prevalent twmporaturo. (,'tl) has been
mombor of its clans when first invented, it confirmed by experiments at pressures less
wtis milled by its inventor the molecular than 0-001 mm. Gnodo (17) has shown that
iininp," U is convenient to retain the term at higher pressures is greater, probably

and confine it to tliis type, although tho owing to the presence of a gas film on tho
(second typo, invented later, has an equal bounding surface.
right to it. Consider a layer of gas between two infinite

parallel plates, distant li from each other, to the auxiliary vacuum (H.P.V.) and nof tho

moving relatively tn tho gn-s with velocities middle groove to the L.l'.V. A IN run at
i, 2. Let j) } , j) 2 bo tho pressures of tho gas about 140 revs, per aco, by a pulley and
at points distant L along tho direction of motor. The axle passes through an oil box
motion. TC tho pressure is HO high that the which seals the interior of the housing from
moan free path is small compared with the tho atmosphere. The intrusion of oil from
distance between the platca, tho forces on tho tho oil box is prevented by an Archimedean
gas nro due to its viscosity ; tho relation screw out on the axle, which driven tho oil
between ;>, mid pa is given by the equation bftukwnj'd ; tin's arrangement makes it of
similar to that of Poissoiiillo ; gmit importances that tlio auxiliary vacuum
should bo turned mi after the pump if) star-tod,
(32) and turned off before ifc stops,
Tho precise cnloiilfttion of tho pressni'CH
But if the distance between tho plates is obtained IB very complicated for there 3ms ;

small compared _\vith tho mean free path, to bo taken into account, besides the driving
tho conception of n viscosity depending on of tho gag from n to m by tho friction of tti
collisions between molecules becomes in-
rotating cylinder, tho leak of tho gua fiaqk
significant, and tho equation must involvo from m
to n past tho obstruction C and over
only f, depending on collisions M-itli tho walls. tho surface of the- cylinder between aiiciicHHivo
It is fouwl tlmt
grooves. But theory shows and oxporimcmt
= 5H + . .
confirms that at a sufficiently low preflmmi
Pi K 1 a ).
tho ratio of initial and final pressures in

Tho ratios of the pressures at opposite ends proportional to tho speed of rotation and
of tho plafcos is n function of tho velocities independent of the pressure, hut tho rntin
and of the geometrical quantities : it is in- falls off when
tho pressure in any part of tiio
dependent of the pressures, If any geometri- pump above that (about 0-02 mm.) nt
cal iU-TftilgiMnent other than that of parallel which 6 equal to tho moan fvco path.
is At
is used, this a speed of MO rovs. per sec. and an auxiliary
plates proposition is still true,
so long as tho pressure is sumciioutly low, vacuum of (H mm. pujpj, is about 1() , HO
and so long as tho volocitica are consider- that a pressure of 10~ mm. can ho obtained.
ably less than the mean voloeity of the But the ratio varies with tho gas, in virtuo
molecules, Jf this last condition WTO not of tho occurrence of pg in (III) ; it is less wiUi
fulfilledtho distribution of velocities among hydrogen than with air; probably hydrngon
tho molecules Jonving Hie boundary would bo
no longer Maxirolliftn, and f would bo greater
which troiild elefti'h/ bo desirable for tho
purpose in viaw.
(-10). Tito construction of Onedo's ])uinp
(18) in- which this principle is applied is shown
(Hug ram in aticif illy liy transverse) arid longi- fj

tudinal sections in fig, 28. A is a cylinder

formed a large part of the gas with which thin

FIG. 28, measurement was obtained. Lower prmmir^H
could l)o obtained with a bettor
rotating in (.lie closely fitting housing B in ; vacuum but tho viwiuum attainable IB doll-

the surface of A arc cut grooves into which

nitoly limited by that of tho auxiliary pinnp.
project the obstructions attached to the An important feature of these pumps IH i,Uo
housing ; tiio pipes and m
open into the groat speed of pumping. Fig, 20 A shown H
grooves on either side of C. If A rotates 3
(in cm. /see.) plotted against tho pimsm-n
clockwise tho friction between tho
rotating (log. scale) ; for comparison, TA abowa K Tni-
cylinder and tho gas lowers tho pressure at tho Gaedo rotary movoury pump (Ah),
n and increases it at m. The grooves are in Tho molecular pump would have wailo
series from tho middle outwards of the ; m possible modern high-vaouum woi-It ; hut Jill
middle groove is connected to n O f each its advantages, except ono, ai'o possomofj
o the grooves on either Hide, and so on ; by tbci next typo of pin|> to bo oonaJilPiw'l.
i of tho outermost grooves ore connected This one advantage is ihak it will remove tl

vapours, while all other high-vacuum pumps Fig, 30 (a) worn adopted, tho diffusion of gas
leave morcury vapour, whioli has to be removed out of .L would ho opposed by a vigorous
by condensation. But the advantage ia of stream of vapour entering ; if the gas from H
iittlo practical importance for most work could not diffuse against tlio stream neither
since tlio pump will maintain the vacuum could tho gas from L ; there would bo no
only while it ia running if it ia to bo stopped
; pumping. Some dovine, therefore, must be
and tho vacuum preserved, some form of tap adopted to prevent a stream of vapour
or trap must ho inserted, and such devices
entering L.
always introduce vapours. On the other (42) GAEDE DIFFUSION PUMP, The device
hand tho molecular pump ia necessarily originally adopted by Gacdo (1!)) was to
expensive- and requires skilled attention. place in the tube loading to L an obstruction
Despite its novelty ami ingenuity it ia already with a, very small opening. If tlio linear

practically obsolete. dimensions of this opening are small compared

with the mean free path of tho molecules, tho
('k. Diffusion 1'umps
laws of tho flow of gas and vapour through
(41 ) DIFFUSION PUMPS. In Fig. 30 () lot tho opening are not those of liydroclynamical
H bo tho 1T.P.V. in which is maintained atroaming, but those of diffusion. Tho flow
a constant pressure, L tho L.I'.V. to 1m depends on the partial pressure of the con-
evacuated. .List X
ho a vessel in which some stituents of the mixture and not on their total
liquid can bo boated, while II, but not L, pressure. Since the partial pressure of tlio

is cooled so as to condense its

vapour. If gus in tho tube XMH
is zero, tlio gas will
tho liquid is heated to a temperature at diffuse oufc through tho entering vapour in
whioli its vapour pressure P is largo compared spite of the fact that tho total pressure of tlio
with pn , a continual stream of vapour will vapour is greater in tho tube than in L.
pour along the tube XMH, driving tho gas The problem can bo treated exactly by
before it and condensing in IT if the stream ; molecular theory. If d is tlio diameter and <r
thc aroa of tlio opening at M, A tho mutui fi'to
path, p tho density of the gas at a piTHHiiro
of 1 dyne per em.*, then tho volume of the gas,
measured at j>i., issuing through JI par see. is
given by
B = a- _- f
. . .

is sufficiently rapid tho gas in II will bo un-

able to diffuse back into tho tube against it.
On tho other hand, tho gas in L will diffuse a attains the maximum 1 when <lj\ is mnall ;
out into tho vapour stream and bo carried but o- decreases with d, Tlio maximum value
by it into H. For this diffusion is not opposed of S, when A is lixod, is given approximately
by a vigorous stream in tho contrary direction ; Ijy <Z
= \. This maximum will increase with
since L is not cooled, tho vapour will not \, which, since tho vapour pressure nf tho
condense in L, and vapour will enter it only liquid muat bo greater than p lt , in limited
at a rate suflioiont to replace tho gas diffusing by pn. Accordingly tlio speed of the pump
out. Accordingly, after some time L will bo depends greatly on tho auxiliary vacuum,
completely evacuated of gas and contain and also on tho tomporaturo of the liquid.
only vapour. If L is now cooled, tho vapour For if P is too amall, gas will (M'unn biusk
will oondonso and an almost perfect vacuum from H if it is too largo, tho dillnsion of

bo loft in L. Tho vacuum will not bo quito gas from L will l>o lihidernd liy Lho onpuH-
perfect because Homo gas from will diffuseH ing flow. Tho conditions in tho pump nocil
back against the stream of vapour, however therefore careful adjustment- On tho other
low is jj|[ and however rapid tho stream ; but hand, S is independent of tho ^IVCHBUIO oE tho
a consideration of tho magnitudes involved, gas and dependent only on its nature and
wilt show that. tho residual pressure could tomporaturo this ia tho most
striking feature
easily bo made inappreciable. of all diffusion pumps. K inCreator for Lho
Such in the principle of the diffusion pump lighter gases tho variation of >S with tho

in its simplest and ideal form. In practice- nature of tho gas is tho contrary of that for
it is impossible ti> maintain the wholo of L tho molecular pump.
{tho apparatus to he evacuated) at or ahova (<I3). Any liquid oould bo used in a, dif-
tho temperature of tho boiling liquid during fusion pump, HO long as it could be miuntuined
the evacuation. L as well as IL is cooled at tho appropriate temperature. Actually
sufliciontly to condense the vapour, and mercury is used, for the appropriate tempera-
consequently if tho simple arrangement of turo is convenient (P^O-U mm. nl. HH)" (1.) ;
its density,
moreover it is chemically stable and does nob vapour stream, as measured by
may bo very much greater than tho presauro
mst glass. But its universal adoption is prob-
association of in L. (It will bo seen that tho construction
ably duo in part, to the previous
based is similar to tho gas injector pump Be,
mercury with air-pumps an association Tho
but the principle of- action fo different.
on quite different properties.
gas from L diffuses against
tho hydrostatic
The vapour of tho liquid used in the pump
not flow with it.)
nt tlio pressure cm-responding to atmospheric pressure ; it does
If tho walls of tho outer tubo wore
heated by
in L. It is
temperature is left by tlio pump on them
connected the vapour, tho liquid condensing
easily removed from
to Jj'by making the connection through a trap
would have a vapour pressure greater than
of vapour
Tho introduction of tho that in L ; there would bo a (low
cooled' in liquid air.
from tho heated walls towards L, whicli is
cooled trap involves, of course, a continual
the diffusion of tho cool, and this stream would once more hindei
stream of
vapour opposing
tho diffusion' of giia from L. Accordingly
g!iathrough L ; but at atmospheric, tempera-
Langmuir who first used tins arrange-
tures tho vapour pressure of mercury is so (20),
ment, laid great stress upon the cooling
low that tho consequent diminution in tho stream
tho walls struck by tho vapour
lit ;

spend of the pump is inappreciable.

insisted that the vapour must ho immediately
Gaedo (19) hits pointed out tliafc tho existence
o this stream causes n slight error, appreciable
condensed to the temperature prevailing ir
L, so that there should
he no flow of tho vopom
ab tho lowest pressures, in tho measurement
tho pressure in Ij by a MoLood gauge. back towards L. On account of tho import
to anco attributed to this condensation, ho tormet
Since vapour ia streaming from the gauge to distill
L, tho pressure of the gas in
L is slightly higher his pump a "condensation" pump but i
diffusion pump
guish it from Gacdo's

than its pressure in the gauge.

is equally a diffusion pump in
tho senso tha
(44). The original diffusion pump of Gactlo of parthi
involved complicated glass construction since : tho gas from L follows tho gradient
that of total pressure. Gehrt
it,is no longer used, it need not
bo shown. pressure, not
(21) has pointed
out that Langmuir's prinoipl
The maximum value of S obtainable was about who did no
80 em. and far below that of tho molecular was anticipated by Magnus (22),

at tlio higher pressures. On the other sco its applications.

pump It appears, moreover, that though tho vor;
hand S was, as theory predicts, independent
to tho limits of measure- efficient cooling of tho walls ami tho complot
of the pressure down
ment at pressures loss than l()-
mm. tho condensation is necessary to tho most oflicion
to make
diffusion was as good ns tho molecular pump, working of tho pump, it is possible
pump of this typo with much
less oflioien
and no practical limit to tho pressure was set
This is achieved in Omvford'
of gas from 11 against tho cooling.
by tho diffusion But in its worldn
parallel jot pump (24).
vapour stream.
ib should bo observed that there
characteristics this pump resembles tho di.
tho pump to fusion rather than tlio condensation pumj
is nothing in tho principle of
Ihnib its use to very low pressures, except tho and has not tho advantages of Langmuir
condition tliab d = \ if openings as small as
: pump noted below.
The construction of Langmuir
tho free path could bo obtained at atmospheric S (46).
it en
pressure the pump
would work. Gaodo lias pump in metal (20) is shown in Fig. !H ;
used at also bo made without great complication i

actually the pump atmospheric pressure,

as tho Tho mercury M is maintained at;
taking tho pouea of an earthenware pot glass (20).
but since toniporatiiro of about
100 0. by tho expend
openings and steam as the vapour ;

the pores aro hacked by very fine tubes, turo of about 300 watts, supplied eloctricall
hofore it or a burner. Tho baffle B deflects tl
through which tho gas has to flow l>y
vapour stream downwards and against
arrives at tho pump, tho speed of snoh a
walls cooled by tho water jacket <T, Tl
is very slow it is not generally of
pump ;

tho auxiliai
L.P.V. ia connected to L;
praetioal use. " " to H. If this pump nrnm.tn.ins
A simpler and more efficacious method of pressure pa of O'Ol mm. or less, S in an gi'tt.
into L is to use as 3000-4000 cm. /sco. and is, as btrfnro, i:

preventing tho flow of vapour

of p tl down to tho lowest obsor
the inertia of tho stream to carry it past the dependent
shown in able pressures. Higher pressures pn clecrea
opening. Thus in the modification
of tho stream of tho speed, but tho pump will work oven
Fig. 30 (/>), if tho velocity
is as groat as the velocity tho pressure is nearly 0-1 mm. Tho spo'i
vapour issuing at of t
is independent of tlio temperature
of tlio moletmlcs in the stream, all tho vapour
will travel forward till it meets tho walls, of mercury, so long as this is abovo a lim
tho outer tube or the gas in none will H ;
which is greater tho greater is p a Tho gre.

stream towards L and prevent tho diffusion speed of the pumpgreater even than t

of gas fi'om TJ, although the pressure in tho maximum of the molecular pump and t
absence of any need for the accurate control cm. 3 /seo., and is sot by Lho dimoiimnr ..... f

of temperature, arc tlio

advantages that have inlet tube" (sen below) rather than l, v
caused condensation pumps to replace
wholly pump itself. Tho auxiliary vacuum i.l I

tho original Gaodo type. mm. for this or tho Liingmuir pump nm
Many variations on tho original obtained
Langmuir by oil -pumps, and jnv:n<iii <

design have been made. Tn some of thorn difficulty. The simple pump of /ify. ;i;',
well as tho Langmuir and other nii>n> .,.
plicated pumps, is used on the lining
scale in thomanufacture of lilninnioii.ii' \n\<.
and other high-vacuum devices.
(47) TECHNKJUI':. li, | (

been pointed out that fchcro miiHl. bit nu

limit to tlio pressure reached
by a ilihu.i.
or condensation pump, determined by "
tho speed of the vapour stream. Win u
high -vacuum pump is usud Lho |iiv
attainable and tho speed of piim|>ini? ,

actually limited by factors other (<hnu il

clliciency of the pump. In tho lln.i, U, (

the tubes connecting the pump to tho

a|>| mi ,ii
offera resistance to tho How of JJRH. Kmnj: .

(10) has shown that tho volumetric) HJH <>i| ..|

pcifcct pump is given by

If Hi, 31. where H is a coiiHtant dojiendciil. nf I In- I-

of the connecting tube and />

w llu- dm
27) tho heating of the mercury is effected

hy an are maintained between two mercury

surfatics inside tho in fact ihia
])uiii]> ;

arrangement lias boon applied to large

mercury- vapour current rcotifio.ra (28), an
that the rectifier acts as its own high-viiciiuni
pump, only an auxiliary pump being needed.
Again, it has boon proponed (2i),30, 31) to
combine in the same- apparatus a mercury-
vapour jet pump and a condensation pump,
using the same stream of meroury vapour.
Tlio two act in series, and the combination
will work with an auxiliary vacuum of 10-20
mm. ; but tho construction SB complicated.
However, special reference need be made
only to one typo of this pump, remarkable for
its simplicity. It is found possible to dispense
altogether with the inner tube in Pig. 30 (b)
and to make the arrangement of Fig, 30 (a)
act as a condensation pump by merely cooling FIG. 32.
the walls of tho horizontal tube. Since the
mercury molecules striking tho cooled walls of tho gas at a pressure of 1
clviir/fiii.' *,

tlo not rebound therefrom, if tho cooled tube a cylindrical tube of radius r and icujidi
' 1.,

is 'made suflieiontly long, all the molecules

emerging from its end at M will bo moving "
parallel to tho length of the tube and will
not outer tho side tube. One form of such a If r is expressed in millimetres, I, in m< n.
then for air at 20",
jiump is described in (32). An oven simpler
construction is shown in Fig, 32 adopted, tho
whole being made of sheet metal. The limit
iif 2 u at which tho
pump will work is about S ia 1000 cm, a /sco. for air (loiviug
0-015 mm., and somewhat higher than that tube 1 metro long and nbout 1 tun. in
for the Langmuir type. Hut tho value at Oonsoquently to make ffull, use. of |.h.< !.j...
which tho maximum speed is obtained is not of a condensation puilip, coniutciliuj 1 in'
very different ; this maximum is about 1500 not less than "2 cm. in diameter miwt ln u. -- >

ifthe evacuated vessel has to ho sealed oft' temperature at which an inexhaustible ovolu.
eventually,tlio speed iii often limited by the tioii of gus starts. Tho gas lima introditml
diameter of tho scaling - off constriction. H can bo diminished by heating tho Healing-till'
lie sealed off by a blow-
this constriction is to place to near its softening torn pora turn fur
pipe it in difficult in make it iiiora titan 3 innt, sonic time before sealing, and then co
in -internal tlijifiiotor; but it ia possible by tho soiling &f> quidtly a-s possible,
heating the tube very unifnrmly ami by much of tho gas (chiefly water vapour) UIUN
malting tlio temperature gradient along it introduced disappears, boing either
" "
very stoop (e.g. by a small electric furnace by tho glass, or cleaned up by a
surrounding it) to seal off tubes 10 mm. or subsequently passed tlirough tho vessel. But
more in diameter. it scorns that, whatever pneciuilifm in tfilton,
(48). Secondly, there is an evolution of tlio most iltjiuato forms of manometer 'ill

gas from tho apparatus being exhausted. always dotcet tho presence of some gas in a
Glass and metals in their ordinary condition vessel immodiatoly after it ia scaled off.
give off large quantities of gas when exposed For furthor information on thcso points
to u vacuum, The gas from glass is chiefly reference is inado to (83), which is tho IHIHI
water and C0 3
which has boon absorbed from summary in English of the. state of tin neb
tho atmosphere and will bo rcabsorbcd if which ia described in patents rather than in
the glass, having once been freed from gas, scientific journals. Some important dovint'H
ia exposed to tho atmosphere once more. are still kept secret.
Tlio gas from motals ia largely hydrogen and
carbon monoxide, absorbed from flame gases I). MISCELLANEOUS METHODS OP
during manufacture and diffusing out from
the interior, Tho evolution ia greatly hastened
by heat and, in motals at least, by making (49). It remains to consider some other
thorn tho electrode of a discharge, oven if it methods of evacuation which, though thoy do
doea not cause- material heating. not satisfy any definition of a pump that.
To obtain a high vacuum, it ia necessary would liiivo Leon acceptable twenty yearo atfo,
to heat the glass while tlio vessel is exhausted satisfj', ideally at least* that gimi (it Urn
to tho highest temperature that the apparatus beginning of this article. Home of thorn arc,
will stand without collapse about half an;
and still more have been, of great praulinil
hour at this temperature will liberate tho gas importance ; and they do not appear to dill'cr
from tlio surface, but there ia a continual more radically from the older conception of a
evolution at this temperature which ia generally pump (winch implied a nioeliftiiioftl dwicn
thought to result from an actual decomposition with moving parts) than tJio diJl'iiaion piiiii[V).
of the glass ; this evolution stops when the Their chief modern UHO lies in the potifiiliHII.y
glass is cooled, but if tho cooling ia too rapid thoy provide of evacuating a portalilo veswel
some of the gas may condense on tho glass soaled off from all fixed apparatus,
during cooling and be evolved slowly again. (00) C'ounnNSA.'1'iON. Tho presaum of tlm
Ifor some purposes it is desirftWo to enclose the gfts in ft vessel can bo reduced by <n'liitf(
apparatus in a, vacuum furnace, so that tins suJiTicionUy any jwirt of it wnlla, Tho limilJit^
external pi'L'ssuro of tlio atmosphere is removed pressure) obtainable by thin means in, of tmimii',
and the glass can bo healed for some time the vapour pressure of tlio Hiibstamto at Urn
beyond tlio softening point without collapse. lowest temperature available, Savory, when
Tho metal parts must also bo heated to near ho evacuated tho cylinder of his H(.oikin-|iump
tnoir molting point line severnl hours ; this by condensing tho steam with ft jot of wnl,r,
heating is effected in modern practice, either was using this prinoiplo. It has bad
by malting tho motal the target of an electron nioro modern applications, e.g, when a
bombardment from on incandescent cathode, such as air has been removed by (linpl
or by oxoiting high-frecruonoy eddy currents in it with C0 a and then condensing UMI
tliometal by coils surrounding tho apparatus, in 'Jiqtiid air. Tho mothod ia also uwd fur
Much time onn bo savntl by beating tlio motals compression. Oltlorino, CO B and fiO t Ituva
in a vacuum before they are introduced into been comprosBcd into containers by cimilt-UMn-
tho apparatus. tion in placo of by compression pumpH.
By long-continued treatment of this nature (51). OIIMMIOAL ACTIOK, Oas nun ahio In*
tlio evolution of gaw can be
stopped and a removed by causing to
react (liiomii'iilly
vacuum obtained which la perfect so lav s with tho formation of solid or liquid com-
the most dfljiuato nwinomotors can to]], and ia " "
pounds. Gns analysis by absorption irillt
maintained indefinitely it the vessel is gas- liquid reagents employ's this priniiiplci, 1m t
tight, But if the exhausted apparatus is aomo dovolopments of it nond moro Hjie^iul
aoalocl oft from the pump, some gas is always montion. Thus it has boon shown (!M) tlni.lv
introduced by this operation. If'or in order oalouim hcatod to 700 0. will
inotallio winMim
to so/ton (.lid glass it must bo lioatot! above tlio i
with moat, if not all, gases, except tliomi uf
tho iimotLvo group, to form solid compounds slower than that of the Langnmir
pump at
with low vapour 'pressures, lint some of lower pressures.
the- oompmmtlg,
especially the hydride, have As regards (2), there is some conflict of
conaidorablo dissociation pressures at evidence which has been only
slightly partially removed
higlior temperatures, and tho toinporaturo of by the very complete study of the absorption
tho mtal must bo carefully controlled. Tho by charcoal which resulted from its use in
method has its uses in special circumstances gas masks during the late war. In general
(e.g. whore high vacua have to bo maintained tho denser charcoal from the harder woods
away from a laboratory or supply of liquid air). shows tho greater absorption ; the shell of
Tho alkali mofcals will also combine with all the cocoanut and the kernels of some fruits
active gases the combination is usually
; are the best raw materials. Tho original
brought abtmt by tho electric discharge. It coking should bo at a temperature not exceed-
has long boon known that a discharge passed ing 900 C., and must be followed by some
with a cathode of sodium or potassium process for the removal of residual hydro-
(more- conveniently the alloy of the two) will carbons. For this purpose heating in a stream
remove, the common gases down to tho of chlorine at 800
followed by heating at tho
pressure wlioro tho discharge ceases. Tho same temperature in hydrogen has been
latest development m tins direction is absorp- suggested ; but the best modern practice
tion by heated thorium or zirconium (35). appears to bo alternate absorption of air or
(52) AnaoiirTiON. But such chemical oxygen at atmospheric pressure and liquid
methods hfwo been littfa practised since tho " "
air temperature with out-gassing of tho
discovery of the powerful absorption for gas absorbed gas by evacuation at 400-500 0.
of charcoal at low temperatures. From the Some writers maintain that all absorbed gases
discovery of the method by Dowar (86) to tho can be removed by heating to 600 C., others
invention of the molecular pump in 1013, that heating to any temperature over (500 im-
it was tho standard method of producing pairs the subsequent absorption. It has been
extreme vacua unattainable by liquid piston found also that charcoal, activated by special
pumps. A glass or, preferably, silica tube processes, will produce high vacua even

containing a. Tew grams of charcoal is attached atmospheric temperature.

to tho vessel to bo evacuated. Tho charcoal Absorption, similar to that of charcoal,
in heated (hiring tho preliminary exhaustion displayed by other finely divided solids. ]
which should bo carried to -001
of tho vessel, probably absorb some gas at a
fact, all solids
mm. tho vessel is then disconnected from tho
; temperatures, tlis differences are merely >

pump ami tho charcoal tube cooled in liquid degree. Of the other solids proposed fi
air. H the vessel is largo and the highest practical evacuation, palladium black (whit h

vacuum IB required, two or more charcoal will absorb other gases as well as hydrogci .;

tubes may ho attached, ono being sealed off and finely divided copper may lie montione I,

before the. next is cooled. In addition reference may bo made to tho m .-

Many experiments have been made on (I) usually great absorption of hydrogen by tni .

tho relative amounts of different gases which talmn at atmospheric pressure. Here aga n

charcoal will absorb at different temperatures reference may be made to (38).

and (2) tho absolute amounts absorbed by (53) ABSORPTION IN THE ELECTRIC Di
charcoal prepared in different ways. A full CHARGE. In the early study of X-rays it wi
discussion of tho results is beyond tho scope of found that a hard tube often became harder
this article, and for fuller information reference by tho passage of the discharge through it ii

may bo made to a good summary in (37). consequence of a disappearance of part of th

As regards (I) it appears that, .in general, residual gas. This disappearance seems to b
gases arc morn absorbed tho higher their boiling a normal accompaniment of tlie discharge
points, the exception being the inactive gases
when it does not take place, or when th
which are but slightly absorbed. The mass contrary process of an evolution of gas occurs
absorbed is proportional to tho mass of the it is because tho normal disappearance i

charcoal ;
increases aa tho temperature is
it obscured by an evolution of gas caused b;
decreased and as tho final pressure of the heating or possibly by some other and distinc ;

action of the discharge.

residual gas is increased ; it is doubtful,
therefore, whether Ji really perfect vacuum The' faots concerning this absorption of ga i

could bo obtained by tho method in ideal are still obscure, and still more obsom'e
conditions, bub, as with tho diffusion pump, explanation ol them, It is certain that
tho actual limit lies beyond the range of inactive gases are in general less absorbed I

measurement. Tho rate of absorption de- others, but whether and to what extent
creases greatly aa tho equilibrium pressure nature of the electrodes and of the \\nna
is attained, and, though tho speed of evacua- determine the absorption is not yet certain.
tion is rapid compared with that of any Here reference will only bo made to those
" "
actions of the discharge in clcaumg-up
piston pump down to -0001 mm., it is probably

u-hieh ivrn of A pusses between

tho oppcmibn ci** ri1

l^an practical iiiipnrl-iuico.

bo givon bo (H9). of tlic iilinnont which net as olcdbrudon (*''
ffonoral rnrnixuico may ;

Tlio fiiml ovarimtion of nin:h appni'atiiH as catlioclu,bDing inuatidosuoiit, gives a thcrminii^' J

llitirmiimio valvc-s, niuliuoi'H mid X-ray tuhes omission sumoieut to abolish tho catiiodn *"-' 4
in pi'obulily (stTuolud by thn disohargo. How- of jjotcntial and permit a discharge tn |)JL:'-
ever mri'fidly tlio ttppaniUis in evacuated by uvon when tho potential differoneo Imtwi*'"
intmdiiced in tho olecitrodes does nt exceed fifty vltn.
pumping, uomo gfiH IH alwayt) : *'

off from tlio pump. '1'hia gaa IB Tn Jltulignani's original method tho p1n
muUiiiK tlr
introduced by hi-fU-iiii^
lai'tfoly, if nob mthuly,
u.l>snrl>nd by tho walla plioniH vapour was K"
Htmill qiiiinbity of red phosphorus in tlio l-u'
ami i-hiiitriJiUm hctoro any disohurfjo passes; >l
bub during tho iirt fow momenta, of bho dis- oonnoeting tlio lamp to the pump jimb bi<f ?*
Tho labor practice is tti til(l
tho normal fwnotion sealing off, d(i|
isharyo, wliidh ropreKimtH
the red phosphorus 011 tho filament or
of thfi npnuvatiiH, Kiiino Jui-Lhoi clmngo occurs

" " inoro tv

adjacent supports, whence
it is oviipornti^l /*
\vliioh imilus.! tho ('.loan-up oomjiloto **
mom soon as the filnmont is heated. Nowiidii,^'^
and pdriniiiujnt. * " '**
been known tlmt tho passage is also usual to depoait on the filament, fcugol-1
Ib huti loiuj; H>l

o( bmi dinnliiirgo IjuUvRirti Hintiiblu eloctrocles with phosphorus, salts such as fiuoridrH
' '*
chlorides. Various benefits are aUi'ilmi<'< 1
wotihl proinobo olunniuul notions which lead '*"
tho preaentso of thcao salts, but
it HceniH
to Uio romoval of gas. An itiHtiuico is provided 1

is subsidiary to
Imtwimn nloctrndcs of tlio agreed that thdh' action l-l^'
liy tho diHohar^n " ***
him been already mon- of tho phosphorus, and that tho evaeiui.1
* ^
11 mislulrt, \vliiuh
it has hoon found that a would not occur unless phosphorus (or <mo
the other olomunts incntionod) were proM^i^l

to (in oldctwido of oharcoiil l
this process of phosphorus ovaouiLi*'
' -

ijiuino Lin; ohitrooiil to nbsoi'li at afcmo- By '

aliHorl) without the USD of high -vacuum pumps in Ij^rni'

boiiipmuturo an it -\vill h^-* 1 i
til liqiud-mr tomporaburo (40).
manufadtiiro ia rendered uimeccHsary.
if the piim]) leaves residual gas at a pi'<*MHir
tho most pracitieally imporb-
S {fi.|),--Uul
on tho of 0-1 nun. in tho lamp, almost all this KUH
iuit uroutiHH of ovaiiinilion dopondonfc " litn
bo ronioved in tho fit-fit few seconds of
in blmb which involves tho introilnti-
ovacuatod ing," and tho prosauro
reduced to ICIHH Uun*
lion of ])hoH]ihonm vapour into tlio
It n.]i!'iu'rt to liavo bom diaoovorod
0-01)1 mm. No morcuiy-pumps are now in<*-l
vcHHi'l. < 'l<i->-*
it wan n]>|Uod to tho in lamp-making; oil-pumps, usually of
lirnt liy Mitllftiituu (-11)1
nvaniiittioii ('[ (ihmLrio iiKiiuuloHCGiil lumps and Ad, aro Hiifficicnfc. 1

iiwd for the HIUHU ]Jiir]iriHB cioiitinu-

The same method has been applied to it'll*'*
At first ib appcara cominoi-dial vacuum apparatus, c,g. roislify **M
oiinly niniio if-!* diHiiovory. *!*
valves of tho old typo without an miunKU-m-*

hiivn lu'H Ihou^hb that tli
action was
oathodo (43), and, moro rarely, to tho mmi<i n
olmmimU, but ib in now known
bo Im dopmulent
bins thormionio typo. H. *' .

on the IIIHIHUKO f Jt diueluu'Ko bliroiiKh


1. Vf. fliicilo, Pints. Heit., 101!!,

xlv. I2:m.
V .\f*-r
a. S. 1'. 'I'liomiwiin, Tha JtevelaiHiifitl <>J
'Tl'ui ox iwli iiiiiiilUioiin wluch dotorralnn tlio
ctiriiilAir-l'inufl, 1888.
rHiii.i'umiuit HH JIHI tlU (jlwouro,
of Mm It. Jl. W. Tnivm. The (tturlu vf (Jaws, 111(11.
Kiamnilly tlmt if a
ilia- .M. T(il>Lor, J)inal. Print. Jmtrn., 1HIW, >'l
hill, il may IH.I I.

mixture of gas \''(\

olmmit IH jiiuwdil UirowKli any tlio

ft! .11, IhiKcn, iricrf. Ann., 1881, xll, -Illfi, ;

iir viiniHiw nonlnlnln plumplionia vnponi'

iin'HHiiw will \w Hiiliinud
mom m]iitlly and do n,
Wr Umll Hum ulwHUl.
it would 1m if tho phoRpluiraa
Tho ^^ H^b UIH (l!,1-
Ilrlt. 1'at.
U. Stoarh



li.A. Itw., 1H/*1 I'

vaiHitu- vmvci
Knaineeriitg, June 24 nnd July H,
10 HtJI .
tho walls
aimi-awil mvn 1m iimtuiwl liy howtiiiB

II . Hie WcWrtse, A. v. IhcrinK, 1UIS,

,.[ Ui vofwtil tii ft tftinpnmbuio
at wluoli rod . Xciiner, JJs famnnotiv-Jilaerohr 180.1, t

I\I. Volmcr, C7!. .He) ., 1010, lii.

Tlio latest theory 13 .

ohoHphorim will oviviicimto.

tho ff w is
H C. .It. Iiint!, I'lte. [fan, 1018.
of tlu. iwtlcm (42) in that doinwitejl 16 A, Kmulfc and U. Warburg, Poga. Amt.,
mi Lliu wnllH and cj<.v*irod with a olv.
10, Ji. KmiiiHdii. viiin. rf. Phi/a., mnn, xxyiii,
tho notion i.

i>r rwil pliiwphin'HH iirmluand by 17 . W. flacilu, Phut., 113, x

Ana. <l.

18 W. A. Ptos., IDI3. N 1 Hi 7
Oawlo, Ami.
tho ru-o volution of tho g(is II) W. d. Phi/s., 1015, xlvl. .ir. ,
(incilf!, Ann. .

]ii'ovonbn 20 T. liiuiKimilr. PAw. Roe.. lOlfl,

v I. K.
" " "
t n Mi" varninh remains j ib is . Oolirta, i!H. /. (A. J*ftiw.. l(1 () - ' n *

Unit tlio. *>x! optional eleofcrionl
o( tho NhciuplioniH vn-pour aro_ i

(ivi ,. [.t in knmvii tiliftt milplnir, iodino, 24. W. W. Crawford, Pints. Set., 1017, x. f.r*T.
tho sumo way as 25. ])rlt, 1'ilt. 108SOO. 1017. ,,
ariiijnto ant in Hinmwlmt ,

20. L. T. JoncsamlTl.A.IliissoU, /'Aw*. Jto'., ilM

in thin inabtoi', x 801
In mv iiwftiwloBoent lump tho- noooaBury

27. Stomcus nnd Hnlslte, Hrlt Put. 1 57 1 1

H. I li t u

cliUu-, I). It. I

'Mm2, I in 7. AMMONIA CoMriiussioN KKFHHIKRATOH. .Sen
.rails, J. A.IH. Viieui. A'oc., l'JI7, ,xx\lx. "
SIBIL J.tfrigrati(ii," (2),
JiU, JU' .Stii]isoii,./."ll'ns/i.jlffH/..S'ci., M)l7,vll..|l7. AN'OUHWS: INVKSTKIATIOKH ON TIL n EXPANHLUN
81. M. Volinor, Hrll. J'.-it. 140313, HIM).
;t-Jl. Wiffltot-n .li[i>t;h'Lt! tin., Jlrit. .Put. 1387'iG, I01.
!1JI. S. JJiinliiiiiii), (tt'iu-ml J'Hri'.trir, 2tftww, 11120, IINJlETt Hl(l II PliKfiSUKJiS. See Tliorilllll
sxill. 078-<WU anil 1021, KNlv. a-l 8-!ifi!i.
:M, KxpiuiBMHi," (18) (ii.).
E-\ Kmidj-, I'nx: Jloti. Woe. A, 11)07, Issvlli. 42S),
:if>. .lirJLiiil) 'I'liomsini- JIousLon (.'o., Jlril. I'nt, ANII.IN, Si'EiJK-'KJ Hi'iAT OK, dutoi'niinod by
U)!):t08, ID Jo. ;

11. Olrillitlm tins ohnitriiuU

3fi.J. J)nwiu-, J'ruc, I)v>i. fioc. A, 101)4, Ixxlv. S3 a.
I'j-of. JC. by
I 1
y?. S, Dii.tluiitui, Gr.nerat Kke.lric Jtcvieiv, Mini, !
nictliod. Soo Ciiliiriiiuilry, KLuulrkui-l
xx Iv. C3-OCL
of," (fi).
:iH. H. J>nshninn, (leneral lllertric Jleeieto, 1921, "
x-\lv. 21-1-248. fjii IJjfAKU. Sat! .Oyiifuuomcturs," (S)
Hi). S. Dmhtniiii, (ie.ntml Klcclric Rctww, 11)21, (ii.).
xxlv. i)(S-f!8. 24ii-2.Kl. "
JO. A, J[. I'fnml, /'Ays, /ert., 1012, xlll. 871.
Anon KS. Sco Hti'iustui-L'S, Stiriigtli
of ," {^7)
41. MiiUaimnL, Itrit, J'nt. 101211/04. AiiuniiinDEAN SCREW. Boo JlydnuilicM,"
'!^, llcacaruli Stall' of tlio O.Ji.O. Ltd., London,
/'/if/, jl/fiff., 1021, xll. G85. T3io roHiilUut. of
13. O. Lodfie tintl otlitM-s, Brit. 1'at. 2GO-I7A/05. AucmiiMHDiis' PiiiKOiri,]!.
tUo proamii'ca uotitig on a body humored
Alft -
1'UMI'S, 'JSl'l'IOIENOY OF. See Air- in a fluid is equal bo tho weight of lluid

imwi|M," (5). diaplficod and nista upwards througli bliii

USED AS centre of gravity of tlmt Iluid.


AiRSoiiuwa PuMi'S. Soo Air-

" "
Aiu TilBiuiOMETJJB, Sec ThormodynamicH," TURK, Sco Calorimctry, tlio QuiuiLiini
H). Thooiy," (d-ii).
]. Vn'
" Tests o Hen
Sco llotiVra, Liquid Level liulieutorn," ILBO in StrongtU MatorialH.
S (Ifi), Vol. III. Elnaliio CoiiFttfUits, l){(t(.Tiuina('i(n of."
AuUMINIUM, A'L'OMH! HllAT Ol^, AT L()\V Nt(!(id J. nl.(Mit Wpring Buluiuicul
: valuoa tof, llwoiniloi-. (01) (ii.).
tiilniliitod. Sue Elccstrkjal
(iUiiriuiotry, Ualby's Optical Hmjtirdcr. {(11) (v.).
AlftlhmlH uf," (11), Talilo VI. Goneral Mothcids tulnpUsd for Ddwign of.

AI.UNJJUM. '.L'Jii) trade name fm* a tubing cum- S (fiO).

]K3(jd of fused, alumina (Al a O a ) witli a, bimi- Konncdy-Ashciroft Jteeordtir. (111) (iv.)

iug of firciiliiy ami naod as a protcoting slujnth Itooeo's .Uooortler. (til.) (iii.).
ff)ia tlitjrniooloniont at temperatures np to
lliolilo Autiigi'upliic and Antnniatio A]i|Mii'ii-
15CO" 0. .Sco Thcrmooouplos," (i) (Hi.). tUH. (00).
OF FLUIDS VNDHIl HlOU PllESSUJlEiS. SOO "ViiLrot Kngitut, Tlio Wdtor-cooldtl," (It)
'' 1

Thcnuii,! Expuiision," (18) (Hi.). (ii.).

AMMONIA, LATENT HEAT oe VAL-OIHSATION vdOADRo's LAW. Al any on toinporaliirn

OP :
computed, by various writers, for and pi'OHSiiro, oiiital volinnoH of dilTtsrout
tlHToront toinperatm'ca, and tabulated, Soo ganoH oonfciiin tlio satiio number of iiioluonIoH.
" J-riitont WhiEo (ixiiot fur ])cffout
" "
(7), Table V.
lloiit," ^notss only, it in

ASUIONIA - ABSORPTION REFRtOBWATINO approximately Iruo of voul gascn. Hfi,

" "
MAOIIINKSI. So llcfrigtratioii," (5). TliormodynuniioH," (0(1),

" and
STEAM METER. Soo Motors for JinJajiomg for I'rimary frloi^imdary Forni'h
jrousuromont of .Stoum," (20), Vol. III. and (JonploH. (10).
" Balancing of .KiMiipniiMilmg JMuHsoo. (1Q).
BALANCING. Seo Engines and 'Prime Jlovors,
IJftliuieing of a, Itutor, (K).
Balancing of."
lifUtmcing (f Yarrow KohUolt Tmiddy
of Driving AVlieels of Locomotive.
Engino. g (11).

Contrifiigul Conplo, (.1).

of l.i'oiu'-oylindor Engines : In-
Couple OIOHIII'O. (ft).
,oliiHinn of Vwlvu-gcui'. (10) (i.).
Dtilby'a Motluxl. (C).
of Frame- lAmsua. (2). DoiUietions from ]?or<jo and C'onplti Polygons.
of Internal Coiulmation Tilnginos. (?)
.Forco Olosiiro. {(!).

of Loootnotivos. Four Masses on Four Arinn along 11 Sluifl-,

Jialaiioing in Practical Case. (4). (7) (6) and (c).

kinds of Bend
Motion of Connecting Hod. (2) (hi.). Description of the various
Tests fur. Metals. See Hrid.
Motion of Mii.su in a Circle at Uniform (Sit),

H2)(10- .
li'orms of Specimen and Methods of Testing.
Motion of Mass in a Straight Lino at Varying See ibid. (31).
Speed Bennett's Construction. (2) (ii.).
Structures, Strength of," (fl).
Primary Balancing. (10) (i.).
gaiter's Method of treating this 'Problem. (8). UENKENE, LATENT HEAT OF KVAI>OKATION OF,
Seen mlary Haluncing. determined by Griffitlis and Marshall. See
(10) (ii.).
Pour Latent Heat," (10).
Special (JoiiHU'iietLon for balancing of
Mfl-sssos. BERNOULLI'S THEOREM. Along any stream
closure lino in a liquid subject only to gravity
Tho o ^uatinn M= cimploma s {>) = constant,
at a depth
force closure p being tho pressure at a jioint
The equation. M= -

(") zbelow tho plane of reference, / tho density,
Throo Ci-tttilw nt 120 cnnnofc bo designed HO and v the velocity.
that Masses inuliially balance. () (7)
HBAT FutuitEa. See Gases, Specific llent
of, at Higli Temperatures," (fi).
11 BLACK BODY, invented by Wien and Lniiniier
l.'Yiation," (S8).
for tho investigation of tho laws of radiation
AHNKH' TAIILB OP SPECIFIC HEATS OF WATER from a uniformly heated enclosure) descrip- :
AT VARIOUS TiwtrioitATUHES. Sco Moehan- tion of modern' form of. See "Radiation,
of Hout,"
iniil Jiltiiiivttluiil. (1). Determination of tho Constanta, etc." I. (2)
Timi'is. Soo Mechanical Equiva- is tho fraction
lent of Moat," (7). Maximum width of blade along its
BAiuifi, 1880, compared gas-thermometers Radius of propeller
with aoeourttiry Btnndards of toinyorature "
Reo Rcsistaneo and PropuMon,"
in Hit) range 000 to 1000 ami rcoognisod Sliip

tho iinjioi'taiUiO oE a uniform temperature ('H).

ahout tho gna - thonn'omotor BLADINO IN STEAM - TUIUIINES, 1^ ORM AND
Eii-FioiENOY OF. Hco "Turl>ino, Develop-
bulb for pm*jKKJS of high - tompcraturo "
ment of the Steam," (3); Stonm-
inon-HnircmonL IIo iiifcroclucod tlic tliormo-
olomont in tho rrtlo of intormocliary lictwccn turbine, Physics of," (Ii).

tho ga-s-thormoiiiotor bulb and tho tempera- BLOWERS, THEORY OF. Seo "Air-pumps,"
ture to bo measured. Son Tomporatiiro, (!)

llonlifiaUoii of Absolute Scale of," (39) BOMB CALOlUMKTli; I iS

METHOD INTRODUCTION. The laboratory moth'od
BJJAMS, .BiiNiimo ov: MAOAUF.AY'S (1)
" o{ determining tho calorifio value of a fuel in
roil HMVJSUAr. LOADS. Soo Struoturos,
to burn a known weight of a carefully dried
Strength of," (10).
: sample in a vessel containing oxygon.
tho temperature rise of tho water in tho oalori-
TIOH. Sea " Struoturos, Strength of," (12).
motor tho heating value of the fuel is com-
BKAU nr, HOOHASI OYOLK. Soo "Engines, puted, taking into account certain
Internal Combustion," winch are described later. AKlunigh tho
Tliorniodyiminios o
calorific value does not givo all tho informa-
and (fil).
(34) " tion desired concerning a particular fuel, or
determine its suitability for a specified purpose,
ftiwo ami Propulsion," (3),
tho heating
BKOQVEUi'Hj, 1803, ooinparedgaa-thormiunotors yet it is generally accepted that
vahio is tho most important property to bo
with Booondury etandards of temperature
" considered in estimating tho value of fuel.
In the range G00 to 1(100.- Sco Tomjioi-o-
Purchasers of largo quantities of coal now
buro, Eoalisation of Absolute Scale of,"
adopt tho heat-unit basis of evaluation, and
(30) <iv.).
REKIiroiSRATINO MAC1IIINH. tho technique of combustion oalorimotry 1ms
t-r, - COIJiMAN
boon so well developed that a skilled operator
Used for cold stores and tho holds of ships.
can average thirty : fivo determinations per day.
Ah is tho working substance used. Sco

" Two types of apparatus are

employed for
RofrlffowiUoii," (4).
In one tho fuel sample is burned
auoli tests.
UffiMD TKST3 !

Bond Tost bayond tho Yiohl-

under normal atmospheric pressure in a calori-
Altocnafcing " " 1
" meter of tho submerged boll typo, whilst
Potnt, Sco Elftstio Constanta, Deter-
mination .of," Soo "Conl Calorimeter."

in tho o tlior the fuel is burned under high Parr 1 has recently designed a bomb of an
pressure in a "bomb" typo of calorimeter. aeid proof base metal alloy which appears
- -

Sonic authorities prefer tho " boll " typo to very promising.
bnmb " because in it tho combustion is Some investigators employ a replaceable
carried out in oxygon at nearly atmospheric lining, but in practice it is found difficult

pressure and, consequently, llio conditions to maintain a

resemble those obtained in steam boiler perfect fit and,
practice. With a "bomb" calorimeter tho consequently,
combustion is nlmost instantaneous and difficulties arise
resembles an explosion in its violence and owing to leak-
rapidity. ago of the pro-
Tho decomposition products of coal vary ducts of com-
somewhat, and ifc is generally fonnd that tho bustion into
results obtained with tho bomb calorimeter tlie space be-
" hind the lining
are slightly higher than those with the bell."
For scientific work, however, tho bomb type where it cor-
rodes the metal
in universally used, since under good work-
ing conditions tho combustion obtained is of the body of
In skilled hands either the bomb.
practically complete.
jrpotliod gives reliable and Tho staff

concordant results fur solid of tho U.S.

fuels, but the
bomb " is Bureau of
the only method applicable Mines have de-
to liquid fuels, veloped a form
A of bomb(see
BOMB CALORIMETER OUT- Pigs, 2 and 3) in which the lid is held in position
FIT. The ofilorimetrio out- by a novel form of sealing device. This con-
a tough steel receiving nut and kick
'"'' C01ia ' 8ts
elements :
f the following sists of
so constructed that less than a one-eighth turn
with the wrench suffices for sealing. circular A
Tho bomb.
gasket of electrician's solder effects tho
(i.) seal.
(ii.) The calorimeter This kicking device is
vessel, stirror, and
an adaptation of the
constant tempera-
principle used in tho
ture jacket.
breech locks of artillery.
The temperature claim for this de-
measuring instrument. ample
sign durability,
In ono
(i.) The Bomb. strength, and facility
of forms of
tho oldest of manipulation.
apparatus tho Further, when the lock
Donkin. the bomb consists of a massive .. wears out a now ono
gun-metal cylinder provided with a cover held can bo substituted
clown by three studs. Tho cover is pro- without tho expense of
vided with a milled -head screw valve for making and gold-plating
regulating tho Inlet of oxygen to the cavity a now shell. Tfie shell
inside tho bomb. The joint between the of the bomb is made of
bomb proper and its cover is effected by Monol. metal which is

moans of a lead washer Inserted in a circular well adapted to gold

groove. The inside of tho cover has a pro- plating.
which registers with this groove Fury has devised a
jecting ring
ivheii it is serowe<l down. Tho bomb is bomb calorimeter in
l-'ia. ;(.

plutort inside with gold

in order to withstand which tho heat devel-
tho corrosive action of tho nitric and sulphuric oped "is shown on an indicator. See article
acids produced l>y the combustion of tho on Oalorimetry, Method of Mixtures,"
fuel. Tho mostsatisfactory form of lining (13) {ii.), "Metallic Block Calorimeters."
is that (if platinum, but nowadays it is not (ii.) The Calorimeter, Slirrer,
and Constant
much lined on tho score o[ expense. Porcelain '
"An Acld-roslatlne Alloy to replace Platinum
enamel is ftlao somotjmes used for lining tho In tho Construction of a Bomb Calorimeter," Journ.
bomb. ft^ Am. Chem, Son., Mov. 1016, xxxvil. 281B-2G22. A
test of the nbovo by E. H. Jesse, Jr., Eighth Int. Cony,
Tho Krookcr .ftypo of bomb has a cover Appl. Chem., 1012, i. 233, 380. , L
, T ,

screwed on ta^ho bomb (see Fig. 1). Tho

"A Convenient Multiple) - unit Calorimeter In-
stallatlon," by J. D. Davis and E. L. Wallace,
bomb is nuuto of atcol and lias a fixed platinum of Mines Techniral Paper, 01, 48 pp., Washington,
lining, while the cover is of bronze. 1018, Abstract in Engineering, Jan. lOj 191*).
'('t',t/if<.mtitrr. Jtir.knt. Tn Mm outfits employed ovapnration in(.n tho spurn lid \M' 4
* '

in c!iumlry (ho (snlorhnotor, stirror, and

lliisi tho cnlorimelor mid its jaoltet.
jiinkiili similar to those employed in
i.u'o Soldorod to tho jiKskot nre two heavy tiivi 1
" '

III^H, by which it ia hold

it-ppiimtiis for Mi'dinnry ealorimntrio oxpori- to tlio vorti<ial lir*
monlH l.y tlio Motlmd of Mixtures (seo T-bai iif tho framo in smsh IL iimnnor \\n

|i> \if*
(!})- mit tho J!i<;l!t ID slide vortioally. Tho jin-1-.*"''
Ab HID U.8. Hiiremt nf Minos 1 a form in Hiipportdd by a heavy helioal brnsa n|nl f
of iippiiriitiifi hit a ItKon developed which is boai'itiR affiiinHt tho bi)ttrnn of tho lanli, :''*'
rH|in<iia.lly mlrtplcwl for combustion oitlorimctry of Hiicti strength that tho jaolvet vhon ulini'f ,*"" 1

(nen A';';/, .(.),

is hold voi'tidal viih itH top slightly ubovo * I*'*
Tho (snloriinnfor ia nmtlo of heavy shoot fliirfnco of tho tank ivator iijvainst mi mljnH*
brass romforoed at tho top and middle by nblo stop lixorl to tlio T-bar.

stirror shaft arrangomont ia alno nlmT*v**
in Tho upper part of tho nluif|. ***
Fig. 4.
which tho driving wheel ia inoimlod omi-ii^i * '*
of a thioU-wnlIod brass tube, into which fl* 1
lower inirt of tlio shaft tolnRoopos, tho laH*'*'
"' ' l

bning jirovidod with a conical piece ((') nc;nu

to tho end which engager n rocoivn al ll* 1 1


lower end of tho shaft tnbo tiFlor Iho maun*' 1

of a ciinifMil fritstion clutoli whon the wlM*'*

calorimeter SH lowered into place. Tho lun *'*

bearing of tho atirror shaft in (Mirricd by (I**"
lidbniokofc (0), which is held to (ho vcrli-' 11 '
T-bar of tho frame by a clamp which poruiH
of raising and lowering tho lid and c|iini|iinM
in any donirod iiosilion.
Tho calorimeter cover has tubular oiitli*! *' 1

(not shown in Fit/, 4) for tho thcrmomrl "''.

Iho stiiTor shaft, and eloistrioal loads, (to (!=.*
Ihe jacket and iln cover may bo (nl ully
immorBcd in the tank wator during an C\JPC* * "
Tn tho equipment of tho Uurenu ni.v MM* ** 1

outfits arc mounted in one thennoHtntii'iili.y

controlled constant tcmpnraLuro bath,
(iii.) The '1'e.mpm-alitre Measurinff hitih-*z
went, With tho majority of bomb oalorimd *-

ontu'la mercury thonno meters arc fimplnv*-*!

for Iho nioasnromont of tlio toinpnruliiro ii' >'
of tho water. Such instruments have I !*-
DnliillM ill' (!;iliirliiii)(iin': A, vortluil section; It.horl- iidvantiigcs of siniplicily, ehoannisHH, IB i *

y.onlnl NiinMim tlirniiHlia'birihnwlng tulmliir allrrcrwfill ; nmddiuto aofsuriicy. Thoy have tint I i !

(J, Hllllllj I'lllhili; I), hdl'l/.OIltlll SOOllon through til'

lltl advantaged of considerable lag and Im-lt

flonsibility nitjiiirod for work of tho liigl
liniss ImtulH
(not) A). A tubular stirrer well prouiHion, Tlio moronry tliormoiiuilor, iu
in MMilmwd In tko oiiloriinotur ns Hhuwii in Ji. over, powHCHHos tho Gorioiift drawback thut
An oloiitvotla in fanUmod to tho btittom <f tlio morcni'y often sticks in tho boro, jmrtlciil
(itiliiriinoloi't but iiiHulatod from it, which with a falling moniscua. This tronbln
nudavi (loiiiitot wiLSi llio bomb plug whan this bo aomowliat allovmtwl by tapping the nl
in phunul in puHiliini. and Homo obwervorH utilise ii miniature c'Jo<
imlin'iitiotor in supported in its jndcot
'L'lio biiKKcr for this purpose,
on ivory Htuds. TJio jaoltot IH a cylin-
Lliroo Tho
thermometer should preferably !M
driiiiilvoaaol provided witli a ortvor (if brass solid stem typo with its scale divided .to -n:!' *.
ground to a \vator -iij^ht lit. Tliis onvor ifl and tho sealo divisions should extond iltm
|)rnvidisd with a thin slioot-braHH wntor soul tu tin; bulb to avoid uncertainty as In t|

(ami /'((/. 4} t\\il tn llio oiwor proper liy moans magnitude of tho emergent column,
of tJinif! Uiiti inHHlutting widn nf ivnry in snuh For work of tho highest precision a ratit
a mnumir Unit whon tho citlonmotor is in plnoo inotrio rosistanco thormomotor in H-J I

anil tlid (itjvcr broup;!it down snugly tlio water CHHcntial, and a description of suitahl" i

HOiil 11 in iiiiutuotwii'h tho surfuco of tlio valor Htrumonts for fcho jmrpoac wilf ho fmiti'l
in tho oalorinuitor, UHIH surviiig us an oiToetual tho article on Kosiatanoo
Davis mill Wallace, toe. eft. (0 <!))

('*) IVEiiTiions OF ooNmroriNu A TICST. used than with other maleriala on account of ilfi

(1.) (iidihmtion of the Apparatus. It is ncces- lower inthini inability.

Iron wii'o i frcquonlly employed Iincanso it bnriis
niu'y to clotuniiino tlio hont capacity of the
hDiiil) and its fittings by experiment, since it instead of only melting, nnd is therefore morn certain
to ignite din sample. Tho heat of foumation of
is rarely
possible to oalciilato this constant iron osido about 1COO of ir* n.
from the sjieciifio heat of tho materials cm-
is oalorii-'B per gnim
Tlio HiicroHL- specimens required about It cm. of ro w
])liiyotl in its construction, There nro two Ileiwin tho corrccli m
wnigliing ],'!3 ing. per metro,
stiuulnrd methods oE effecting this oalibration :
for the In-lit, liberated in ita eomlmstion amoiml 'it
(a) liy an electrical method baaed on tlio tn about 2 calorics per centime-Ire. Nuphl.hah' HI

input into the calori metric system of a- known ignited readily with 1 em. of wire.
amount of boat measured as electrical energy. Ho aldo corrected for tho small iimonnt of nitrio
]iy burning substances of known heating
acid formed from tho nitrogen eontninedin tlio osygcu.
value. The amount i nciirly proportional to t!io lient
liberated in tho combustion and to tho porcrntiifto
() Tho electrical method is capable of
of nitrogen present, and was determined by tilralion
considerable, accuracy, sinco electrical energy
can bo after caoh combustion. The heal of fwritmlion if
measured with high preoiMion. In
IIN0 3 from N+O+'H.O IH about 231) cilleries per
practice, however, tho method suiVors under
tho disadvantage that it involves elaborate TAIIJ.B I
equipment and is ti mo-consuming. Further,
it is by no means
easy to reproduce with it the
Hatno conditions as
prevail during a combus-
tion test. Tlio electrical mothod is generally
adopted in standardising laboratories, but
for the purpose of a works
laboratory tho
second method is to bo preferred.
(6) To en. libra to a calorimeter by means of
standard siilwtanccs, such us naphthalene or
hcMKoio acid, it is advisable to make about
half-a-dozen combustions, sullicient amounts
of tho standard being used to
produce about
tho average- toinpomturn riso obtained in testa
with coal. Tins method, besides boing simple
and easy ol application, tends to niinimiso
errors such as thoso duo to thormoniolcr
calibration, cooling correction, heat input
from stirring, ote. Dickinson ' has recently
mlotennmcd tlio heat combustions of tho
following sulifitiinoea; naphthalene, benwrio
acid, and sucrose or cano sugar, with a
view to thoir adoption aa standards in
calibration work. His results are summarised Tlio tliitu In tlio nbovo faiblo nro oxprefaeil In terms
in Table I. together with those of previous of tlio ID" calorie.
It appears from a comparison of tho values gram of noid. AH the oxygen employed oniitninctt
from 0'3 to O'D per cent of nitrogen, tlio oofrcotuni
given by different observers for tho same snb-
Ktaiuio that bonzoio acid is the moat suitable-
to bo applied for tho beabof formation of a wna HNO
usually about 1 part in 1000.
in view of tlio close agreement of tho results
obtained. (4) PREl'AKATION Off THE TllST
(i.)Solid Fuels, It is necosHary to oonvort
.Dickinson (omul naphthalene to l>o a convenient tho coat into a small Imquotto or tablnicl fen-
material to work with, lint onro was iieci'sanry in tho purjioscs of test. If tlio attempt in iniuio
bundling Hitioo a gram briquette would lose about to employ tho aamplo in powdoi form tho forco 1

milligram in weight pur hour by sublimation. of tho explosion usually blo\vH ecimo tif it
iSuoroso did not BOOTH HO wotl aiUiptud tin botixoio
iioiclfor Htiuulardisaiioii It Iwa a against tho interim 1 walla of tho bomb find it
smaller heat of combustion and frequently fnila to escapes combustion.
Bituminous fuels us a rule-will form briquolten
Dioldnsim migyeals Unit tho higher rcsulls given by pressure alono. In on-scs ivhoro iiiauiiiciont
Ity ntlior observers fur mieroso may bo duo to tlio fuot tarry matter is present in tlie natiirnl fitol,
Unit thoy may not have- corrected for tlio heat just sufficient of a 1 per cent solution of gum
generated in the firing wire before the snni|)Io igniloH, iiriibio may bo used to make tho
pai'ticloa of:
With BiicroFto n. greator length of fnso wire hns to bo fuel adhesive, For half a gram of fuel, thrco
drops of such solution arc enough. Tho
*[. 2G3, Tiic Times Jlngiiteering iS'iyrfw(. if ?b< aa. 1917.

briquettes must bo heated

in an air bath to platinum wire usod for ignition
110 C. for at least four hours to expel tho generally be found fused owing to the toin |i
last traces of moisture thus introduced before
tiiro momentarily attained. In order l<>
teot tho platinum cnpsiilo 01- crucible from
testing in tho calorimeter.
Fiteh.ln weighing and trans- same effect, and from tho action of the nui
(ii.) Liquid
ferring tho liquid fuel to
tlio bomb, carriers slag produced, it in necessary to lino H ""

blocks of pure thin asbestos board, cift and shaped to fit

consisting of small cylindrical
cellulose are used, one of these blocks being cnicililo or capsule. This asbestos board i'
able to absorb several times its own weight bo dried and ignited before USD in ordo:
The saturated block-, remove all matter that might vithtto
of any ordinary fuel oil.
under the
being weighed, is burned
after results.

conditions and with all the precautions neces- (G) CALOKIMKTBY m r

difference being SODIUM PEKOXITIR. Fusion with wnd
sary for solid fuel, tho only
that a rather higher pressure is lined, in order peroxide is tho only way known for liii'

to obtain a greater supply of oxygen gas in the heat of oxidation of elements

tho bomb. A blank teat with the cellulose not burn in oxygon and which form
alone the necessary data for tho calcula- insoluble in acids, Tho method i
tions. As liquid fuels contain only traces of to tho determination of tho bout formn
acid -form ing elements no trouble arises from of tho oxides of a metal and also tlio
corrosion, -and a bomb provided with a gold combination of metallic oxides, with Hi)*]
lining will last
for somo hundreds of tests. oxido.
As regards tho special precautions necessary Tho method is indirect ami tho heat o
to obtain correct results when testing liquid sought is not tho observed offeiit ;

fuels, it must bo pointed out that tho absorbent burning in Compressed oxygon is
cellulose blocks sold for this purpose absorb where possible. For examplo, whim oiu
moisture as well as oil, and that it is
necessary is burned with sodium peroxide tho oliHi'i

to dry thorn before use for ono or two hours heat (a;) is tho result of tlio following roiuit
in tlio air-bath at 100 0. When saturated 2Na a O a + = Nii CO + Na 3 0,
B s
with heavy oils of high boiling-point they are
to and it is and * equals the heat of formation of ciiu
also somewhat difficult ignite,
dioxide plus tho bout of combination of <mi
advisable to place a little of tlio dry un-
dioxide with sodium oxide, and lows Mm
saturatcd cellulose- in a loose condition around
tho platinum ignition wire in order to avoid required to separate two atoms of ox;
failure of the test from this cause. As tho from two molecules of sodium poroxido I j

cellulose blocks are largo in proportion to their IB = + 20 -h (Na a O + C0 - (2Na,0 a)

!- SO
weight and absorbent capacity, a larger BO that
than for the
platinum dish will bo required
tests with solid fuel, and tho platinum ignition
Moreover, many substances do not givo

wire should bo arranged to hold down tho

cellulose block lest tho explosive violence- of
sodium peroxido sufficient heat to fno
the combustion blows it out of tho dish. mixture), and honoo some readily comlitiH

QUANTITY air OXYGEN iiEQuiniii). substance, sueli as sulphur or carbon, J

bo added which gives in many cases tho \f
Tlio quantity of oxygon required is about
tlireotimes that which will unite with the part of tho totnl beat effect, I'rnfranor Mia
of Yalo University, who has made an oxlc*i
complete combustion. Dickin-
oliavgo to givo
son found that, when the amount of oxygon study of this method, gives tho following m
was much less than two and a half times that obtained, by fusion with sodium poroxid(
comparison with those- by combimUoi
required to unite with tho combustible charge,
there were often cases of incomplete combustion oxygon. They are :

as indicated by a reduction in tho total heat

liberated, as won as by the occasional presence
of a slight amount of soot and by tho odour
of the products of combustion.

Since tho usual pressure employed in routine

tests is 20 to 27 atmospheres, or 300 Ib. to
400 Ib., it is advisable to have a small back-

pressure valve inserted in tho milled -bead

screw in the bomb cover in order to avoid
a groat loss of gas when disconnecting the
oxygon supply pipe and gauge from tho bomb,
after filling tho latter with oxygon. At those
high pressures tho combustion of tho coal
is praotically instantaneous,
and tho thin

found in the sodium peroxide niotliocl i tlmfc HUPPLKMKNTAHY

tiur lea i:lmloura do combustion et do formation
the carbon and peroxide wore mixed in d
dcs ciirbtirc3 d'hydrogtoio Bolides, ct clmleur de
mortar, thus allowing tho peroxide to absorb combustion et do formation des snores, hydrates do
a little moisture which added to tbo heat of rarbono ct aleoola polyntomiques congcnt'rcs,"
Bortlielot ofc Vlelllo, Ann. CImn. 1'liys. (li), 1887, x.
the fusion. Tho value 2(i7-5 for ,'H \> + 40 ia I

lip. 483 anil 45G,

derived from the results of fusions of iron, "
Uber die Methods dor Vcrbrenmuig orfainlBClior
ferrous oxide, fordo oxide, and the mineral Subsliui/en In Hauorstoll' liol liohnn Drurku,"
SLiilnniHui, Klnber und Liinsboin, J. 1'rnkl, C'lmn,,
magnetite with sodium peroxide, and 205-2 .1881), xxxlx. 5011.
was the result of burning iron in oxygen. "
tJbor don Wilrnio worth von CnrboiiBfiiiron dcr
Sodium peroxide absorbs water rapidly from ni'omatisehcn Hciho," Stohmunn, Klcbor und
the Lsingboln, J. Prukl. Cficm., 1880, xl, 128.
air, consequently it should bo exposed "
Oiler den Wtirme worth dcr Oxnlsaurerellio,"
as littlo aa possible, as the hydrated peroxide
Stohniann, KLobcr mid Langbetu. J. I'rakl. G/ietn,,
will give more heat with a combustible than 1881), xl. l!02.
the One two samples, that
of liber den Wiirmowcrth dcs Mcthylalliohola und
anhydrous. fester Mcthylilter," tftohiiiaini, Klebur und Lanqbcln,
which gives off tho oxygen when fused, is
less J, Prakl, Chcrn., 1880, xl. 341.
the bettor one. The error from tho water "
ttber den Wflrmcworth von Kohlen hydra ton,
content is small in good peroxide. Its effect is mehrsiiurlKoii Alltoholcn und I.'lienolen," Slolmiiuiii,
J. J'f<t/;t. Chan,, 1H!>2, xlv. 30ii.
further diminished when carbon, for example, "
Vbor den Wiirmoworth Isomcrcr Siiurcn von dcr
is added to make a mixture fuse, because in Sinsuin monso tziHiH C,1T 0, mid C! B II a Oj," Htolmmmi
practice tho hoat effect of tho carbon is und Liuigboln, J, Pmkt, Cltem., 1801, 1. 388.
Chalcurs do combustion," Bcrthclot ot Lngluln,
found for the carbon and peroxide actually Ann. Ghlm. Phys. (0), 1888, xlll. 321.
usod. Hur In niosnro dcs chaloiiM do combustion, ot
chaloiirs do combustion do divers composite organl-
Various substances may bo added to a per-
(inea," Jiertliclut ot Heoonrn, Ann. Chim. Phys. (0),
oxide mixture to increase tho temperature of 1888. xlil. aa, if()-l.
tho fusion. Mix tor has usod acetylene carbon, iJcacripUoii nf a Boinb-Cnlorlmctor and Method
of its Use," Alwator and Bncll, J. Am, Client. .S'oc.,
sulphur, and lamp-blaok. Pure rhombohedral IDua. xxv. Oiii).
in fine powder would appear to be tho "
sulphur (!bcr die Vorbreiiiiimgswurmo elnlgcr orgnnlscher

best of the threw, but it becomes electrified VcrbiiiduiiBOii," Fladior und F. \S redo, Silzltcr. <l.
Jicrl. Ai-iid.,1001, xx. 087.
when shaken in tho bomb with tho other "
C'oiiocriilnH Urn Adlnbatic Dolermlnatlon of tho
ingredients and sometimes sticks to the bomb Heats ot Comlmstlon of Orimnlo Hubatnnecfl, espoci-
and is not completely oxidised. Sulphide nllySugar und BoiiKol," Mlclinrds. llomleraon and
Am. Acaii. Prof., lOOfi, sin. 673.
is formed, and oeea-sionally free sulphur is Unto (if Combust Ion and Pressure developed in
left. When tho bomb is much blackened by a a Calorlmolric Ifumb," .licnodlct and I'lctclier, J. Am.
Cfiim. Son., 1007, xxlx. 7UO.
fusion with sulphur tho heat result is low. "
Obcr dlo BcHtlmnmiiK dor VerbrommiiKHwHrmo
Acetylene carbon is tho ideal substance to use, orBfinlsclusr Verblndunfiuii mil iUiiiutxmiR des
but difficult to obtain, Pliitiiiwldorstfliidthevinomcters," I'lHtlutr und Wredo,
Us. I'luis. C'ltem., l!)0!l, Ixlx. 218.
One part of tho carbon requires 13 parts of "
J'jielmng dca Variircimungskalorlmctcra unit
pure sodium peroxide for combustion, and it Arbeltwolae,"
.llnUi. Lieb. Am., 1111(1, iscchxHl. 210.

is best to take about 20 parts in determining

Sur In chalfitir iln conibiiHtion ilo quclquea (Mrlvds
]iydronanlitl)ii]6i Unies," Loroux, CM., 101.0, ell, 38-1.
the heat effect of the carbon or lamp-blaok. "An Adiabatlo Calorimeter Cor tlan with tho
For tho combustion of sulphur double the Cftlnrlmctrlo Eomb," lienedlct and Illgghm. J. Am.
C'hcm. Hoc., 1010, xxxll. -101.
calculated amount of poroxido should bo used. " (Jbor dlo
Ticatiimmniff von Verbronnn lifts wit rm on
Oxygon is often evolved in a combustion from ml t tola dcr kalorimetrlsclion Bombo untor JicnutKimg
tho action of an acidic oxido on tho sodium dca I'latinwldorHtandthermomoters," Wrodo, Zeit,
1'hl/s, Olicm., 1010, Ixxv. 81.
peroxide, and the heat required to sot it free "
Jixporlmonta on a Bomb Calorimeter," AHnut,
from tho poroxido IH added to tho observed Engineering,
1010. xc, 755.
Tniluenra of Variation in ftiiecluc Hoat of Water
hoat. This correction, 1-73 g.-cal. for 1 o.c. on Ofilorlinclry of Ifiiela." Loeb, ./. Ind. ling. Cliem,,
of oxygen at and 700 mm., is derived from 1011, 111. 175.
IJekotoff's Na a = 100'2G Cal. and do For-
+0 A Now Method of iKnillon for Bomb Cnlori-
mntcrs." Iloolirlnh, 1'Hglith Int. Cong, Appl. Cham,,
crand's Na a + 20=110-8 Cal.
1012, x. 200.
Apjmratita employed for t7ie Tcsl-s, Tho
bomb is made of sterling silver whilst tho top BOURDON GAUGES, CALIBRATION, ADJUSTMENT,
and fittings are of brass. AND COBIUiOTION FOR TffiMMRATURH OF.
Tho mixture under lost is contained in a "
SCD Prcssuro, Moasuromont of," g (11).
cup of fine silver supported in tho bomb by BOYLE'S LAW on the variation of nronsiiro with
its upper edge. A fusion in the oup cools volume for a constant mass of gas states that
more slowly than when in contact with tho #v=constant (at constant tomporatura).
cold bomb, and hence tho reaction is more Seo "Thermal .Expansion," (14) (ii.)j
" '

complete. Tliorinodynamics," g (5), (50), (00) ;

Tho general arrangement of tho apparatus "Engines, -Thermodynamics of Internal

isidentical with that employed in fuel calori-
Combustion, (13).
metry with the addition of bulbs for collecting 1
Tho jibovo refoi'ciinca luivo boon SRlcRtod from a
any oxygen sot free by fusion. E Q. moro extonslve list quoted by Blcltlnaon,
.AIM-) HuidU.H,: KllKJTIONAl. lUlSIHTAXHt: OK BltlTIBH TllKHMAI, UNIT (H.T.U.). TllO t|1l<H l
" to raino thn loin|)oral.iH'<*
HHAKI; Ih.umtH. Men I'YidUmi," (115), tity of lioal required " "
of'] 11>. of water 1" Kulir. Sen lloal, Co"
A Vl'i IN T VIMJ Ul!' NTH UN A I, ( !()M lUI.S'l'CON

" diiodomif/' "T.honiiodynainicti," S(^)-

MNUINI:. Hen lOnjiiiifn, 'rii(n'iii<idyitJi.nii(:H (2) ;
of liili'iciul t 'iiiiilninlioti,
S (-'') Il1lt '
(' '-^)
1 :il.T. 1-1 . STI-;AM "M HTHJI. Him HIM M1 fi

" (if Stimm," (IN), Vol. 111.

WM I'li'.i,. Hun UydnuilinH," 111.
, Diui'TiNO,
.bYANomi, AND l'Y,\rn:N
.liiuinii;, H'l'HtiNiri'ir in 1

" " nf/
fcioo Hlfiatio Constiiiits, .DoUtrnuiiadiiiu
THIS iii''. Hi'n iSLnioUircfi, WU'on^tli nf,"

t!i< ooul-raofioii \\ lii'' 1
, A'l
VAltlOUH by observing ^
1'roin iw to wiiU **
ri',nil.'i;i!A'l'iniilH liilinliiliHl, \viUi tlio iitcimiis tivkiwt|ilticom tlioohango

MothodH |iro(hi(it!cl by
thu liont yivnii by Inul^"- Um
licjU. Wri' CuUiriiiH'try, KlcKnUimil "
'I'lui (iliscii'vod eliaiigo in iumviTlt'
volnmo 1

nl'," S (HI), Tivliln V.

by uHSiiming a vultut for tin

MAUIIIWIIIN I'lHfHi'ii.TtoKrt Ki'iiiMSMAiiv m 'mn into (saltn-ics

I'AI'll.l.AllV I'lIUK or niiiHs of inorniiry drawn inln Uu> inn|,n_i
' 1 '*

" incLiL tlio addition of nnn nu'im ndi't

A 'rHLiiMdMCTKlt. Hi'ii 'I'lmrnioiimM-y," liy
" JlnMnn ''*
o 't. HDD (Jalm'imiil.riu
S (II) (")
(!AU,I:SII.MI. iiitiltn- in i7 "f pULimnn
n th Ohango of HlnLn," (i!).

llH'i'iinniH'li'i', Mm rcniHliu! of I.w,ModifloafciimH<f. St!(n7u'-/.{i^l >

nf wiv o "Conl CalorinmliiT."

u jnu-t-iiMiliu n|ioi'inu'ii |)hU'imim

of llimsoix'fl Ten: t!m IIIHSKI
di-lcriiiiniul all var'nuiH
lirliif,; cSlrt-i'ily 1
Kon " di'Rwn into Uio iiiHl-nuncnl. by (-1*

lii-viiinvi-H 11(1'
hi Uim"(J. mwuury
nddition of ono moan imlorin nf ln'itl
'.L'luil'imHHiikiLH," () vnluoa fin mm lU'inod and l.abiilnlt'd. S*"-
ll) till)
<!AIJ,r.MlMt'M KliirATION, !1,]lplU)(l "
CiUorimotrio Mollnula IwiMi-d on ll*
'iini tif Htciun uml ;
of Ktatu," Tallin I,
IT OJmng (S),

Liquid Air and llydroj;rn

liaHtiil on im aniihiK'Hitt pri
]<:x|ian> csijil)'
to Uio stoam oaloriiiinl^i' in wliu'l*
1 '
it. -i
iiiH, (Hi) ; ono of tho liqniifii'tl nam'H in iiiii|)loyi-il

oaloriiiititriB substiinflo. H<-(i iliiil, (H).

i .

OAU.MNHAII AMI HAIINKH' MKTIUID ov DKTHU- DllTnronLial Steam, for 1-lic (titlorinlimt-l*

MININU MmiHANHtAl, KliUlVALUNT nV llKAT. of HpLtoifio heats of KIWI-H tit ctuilJHtt

" of iltnvt," volume. Sco rt/d. g (fi).
Hen AUidlmiiit'iil MipiiviUmiL
GHH. Hoo "Gas Oaloriiimtiir."
(Ill OX-
Jnly'H Htnani nn matt'iimi'iil in wliii-b tin'

* P
Hltuun iinrl itylmdor liouli noiifssary to raiwi tho tciupi'i'al.n*
flHiii^'n "f INVU. liolwi-cii
will {Min. /''. /.<'. '--'.A'.. H "7, iisxxi.). I nf a liody from tho air torLiin'i'iitHnt

100" m ini)auroil by dH.nrmmiiitf t-I**'

AM.CNIIAU AMI SWANN. DuLfntiiiiHtiini, liy woMht i>f atoam whioli nnmt bn (Kimlcnri*--* i

Mi" iniiil-iuunim llmv nu'Uniil, r.f Mm Hiuiili into water at 100 to minply HUH lu-at-t .
lH .f air uiul unrlHin dioxide, at. .tn\(i- Hoo Calorimoti'io MnUiodti Imm-d on I tin
ul- air.tt. IHH! 100" <i Sou Cluuigo of State," ('().

liM^rit'iil MotliodHol," (18). Liquid. Hydrogon.

Km i7nV. (K).


Li(tul Oxygon. Sco iftM. (7).
of HonA," (I). Motallin Blook Typ (!il ()ti Mi 11 t '

JTothod of MixturoB," (i:t).

<i[ luUn'iiii-l tl

{} (UK), 'hvlilw U., .111., IV., ftiul V. OALOIlIMliTHA'


Hut incimin-nincnb of hont OATiOiUMKTftY ia ctmocmcd with tlui nutiumt **

All iippm-iit'iiu ftn-
inont of energy in tho form of lifiit. 1
ITin*il in nxtHTlnii'iiUi I'.V
M> (MinBtitutoa ono of tlio most illlllnnlt Imin^lM-
ji, in which
urn mixed. S(in "<'alnn of oxaob monBtiromonts owiiiK to L-' " fl|t1 tttM, 1 '!


AUxinnV a prfot;t jiim-comlnotorof licatdot'H not o^i-n

nif|.r\% MiiMHKl nf ft (R) (I.)-

ttii'rt In* : an iiinininicnt i wldoh UK Tho common method of nnwuirintf mmul i 1 1*

11 "" of luitiii in liy iitiliiiif!; the diifonvnt
In-ill; Hlv.'U mil. liy 11 Iwily tHmUnK^i
in on materials such a.* tlio t'

iwmiionituro ta
<.). li
wniut Itltftior

temperature or the change of state, but in Cootie for tho determination of the specific
recent years another method has conic into float of mercury is shown diagram matieally in
extensive IIHO, known as tho Electrical Method. Fir/. 1. The calorimeter differs from that
In this a definite and easily measurable amount employed for the determination of J in that
energy in L-onvorted into heat, and
of electrical the (lowing mercury is the conductor in which
the resulting change of temperature or state heat is generated electrically and not a fine
observed. Tho electrical niothud has many ad- platinum wire stretched along tho axis ot the
vantages whon measurements of the highest pre- tube as in the case of the J apparatus. A
cision have to bo mado on account of the facility steady stream of mercury flows through the
with which tho heat supply can lie controlled. fine capillary tube and is heated by a carefully
In tho brief review given in the follow- controlled electric current. The difference of
ing pages tho appliances employed in heat temperature between the inflow and outflow
measurements will bo described, and then tho is observed by means of a differential pair of
theories which have been advanced to correlate platinum thermometers. Tho inflow and out-
the thermal data with other physical constants. flow tubes AB and CD are exactly similar,
about 2 cm. internal diameter and 25 cm.

ELECTRICAL long. They are connected by the fine flow

CALOBIMIflTllY, tube BO of 1 mm. in bore and 1 metre in
METHODS OF length, coiled up in the form of a short spiral
(1) GrENEBAr,. The- elcoti'ical method of 2-u cm. in diameter. Tlio inflow and out-
flow tubes are provided with two side tubes,
calodmotry -was first employed by Joulo with
a view to thei determination of tho mechanical one pair for convoying tho current, and the
equivalent of heat (J). Subsequent work by other pair for the mercury flow.
Professors E. H. Griffiths, Schuster, Gannon, A practical advantage possessed by tlio continuous
Callondar, and .Barnes showed that the niotliod flow method in the fact thai tho heat loss from tho
was one capable of tho highest precision for tho walls can ho determined Ly making experiments
determination of J. Tho article on the deter-
mination of tlio mechanical equivalent 1 of
heat should bo consulted for details of tho
method aa applied to the do termination of Iron
'""' Exhaust
Inflow Outflow
heat capacity of water and its variation with
I'lO. 1.
In pussing itmight bo mentioned thai tlio on libra- with different rates of flow, but keeping tho rise of
lion of bomb calorimeters is frequently carried out
temperature constant.
by oleotrieiil methods which un equivalent amount
of heat to that obtained in combustion is generated (i.)
Methods of determining tlie True Mean
in the bomb and its amount m ensured by observations TumpEi-alure Outflow.
of By far the most
of tho watts diMj)ateil, the procedure being identical method is
important practical detail in this
with that followed in methods for determining J, the device adopted for obtaining the true
In specific heat determinations the great mean temperature of tho outflowing liquid.
convenience possessed by the electrical method If a thermometer wore merely inserted in tho
lies in tho faofc that it permits of the deter- outflow tube, leaving a .free space all round
mination of true specific heats, i.e. tho specific for the circulation of tho liquid, it is evident
beat over a very narrow range oi temperature, that the heated liquid would tend to (low in
and consequently it has been of immense a stream along the top of tho outflow tube,
service in determinations of tho variation of and that the thermometer might indicate a
atomic heats with temperature. temperature which had little or no relation to
HEAT OF LIQUIDS HY ELECJ- the mean temperature of tho stream. It is
(2) Si'EOirra
TRIOAI* METHODS. It is obvious that any of easy to malto an error of 20 per cent in this
the appliances which have been devised for manner, A fairly uniform distribution of tho
tlioevaluation of J are also applicable for tho flow might be secured by malting tho space
determination of tho specific heat of liquids, between tho thermometer and tho outflow
and further that they would give data of the tube very narrow. But this leads to another
There are, how- difficulty in tho case of mercury. Aa the
highest order of accuracy.
over, certain dilficullics in practice. space is narrowed tho electrical resistance is
Both Callondar and Griffiths applied their increased, and an appreciable quantity of heat,
electrical methods for thia purpose the former :
which cannot bo accurately estimated, is gener-
determined tlio specific heat of mercury and the ated in tho vicinity of tho thermometers,
latter that of aniline. Tlio difficulty was overcome in tho mercury
(3) SPECIFIC HEAT OF MEHODIIY. Tho experiments by fitting tho inflow and outflow
3 tubes with soft iron cylinders, em. long,
apparatus employed by Callcndar, Barnes, and
turned to fit the tubes and bored to fit the
Sen " Heat, Jlcimaiiienl Equivalent of."
Phil. Trans, A, 1902 7%s, Sev., 100?, xv.
; .
thermometers. The soft iron had a eondue-
l-ivity about ton times Unit of moroury for l. lint if thmi is any HJ-MU-I ' J

V ( (
^s l
both lienl and ehiiitvicity. 'I'ho honk jfoiienited in the toniponiliini <listril)utinti
by tlid diimmt in Um immtuliati) vicinity of cliangn (if How, thon thi^ro nnmt lio n i" f
tlio flicriuometni' hidliH wan HI> Hinnll that H|)ondinj- HyalcniatUs HUTorc-imo in (In- ''*V*
Ihn wnM-M might, fairly bn rulculalwl from fho IONH, wlUisli will lead to ciitiHlant crrorii in 1'
iliftW<!iii!uof potential hntwi-oti HID iron bluel\H ciilmilatioii if no iicwiiint ifi l:nlni of it - ..

Tho of crrnr of tins typn in l'*

"' ' '

n(< tlio nil Idln pi in tn

i of the hi! MM. i
lioHsihlo Hinii'Cio ;

immiury Mlnnim wan fmwd to ciixmlnlo in n In-ill,

hy tidiidiiiition along Uio oiitllnw (i **".''''
li'l 111
npini.1 Miirow thread of miilnMo diiiieiisiioiiH out Wluiii 'l.hn flow in hli'tfo, liculril
in tlio initnr Hiirfacn uf the hliuikEi, whioh piiHftinjr aloiiH Hie. 1-idin will UtH!p il
HO Unit Um j;nn I i 1 '**
provriited ttni formation uf Hfream-liiioH alnn^ a iniiffirni loinjic.raliirn,
' '*'
ntio nil In nf 1,1 lit tuho, and Rroured uniformity in tilts ontllinv lidio will bo Hiniill HIH I

-'^ '*
of toniiKunlitim tb roi i(^l 01 it t,|io iirnHft-Htiution of 1 iiondutition IOHH (!oiTon]i(mdin#ly initnid-.
" l jl "

MHI ontllow tulii', Tim hit-Si iiuiiduittivity (if Iho ilmv JH iliininiHluul, sniiposiiiK tint lcni| "

* ****
(lui iron (lino iiHMiHkul in MCHiiiiin^ tho muno tu of Iho outflow to Htmuin tlio Hiun*-.
l **
I'df-mll, gradient in tlio oiiUhm'-tnlio numt iiunvn^'^ "
A N of tho lluw, t*i''
]-m-n!ni'ly anil Infill ilm'inn fur iivcm^inj; ilin
pro|iorfion to tlio n!i)iproi<nl
tho mdiiition IOHH i-oinuins Hourly tlm n"> |1|T
uiillluw t(iiii]ii'i'jitinii
A van iLjiplii'il in Clio wnlci'
Tim of Ilio tlii'i'inonn-d 1
\vuti will * '**"
I'lilin'inuitri', linlli ]'
Tlio mindiittlicm IOHH vary diroc-tly iif"
nf biuli t'oiHbiulivH.y, on *"**
Hi-adiont, or invorHoly us thn llcw, for u
lilli'il ivitli JL (n]i|ii'i' wli'Dvi >?**
till) Dlltttiijo (if wlliull IL rillilirr U'UH ivninul (o
fijiil'a! i-ist) of
111 Um unUlow tiilttt us ulii'si'ly n ]i<^Hil)lc. Tim
A Hiiiallifrnirnf thin hind, duo to wmdiif* i .

iiimiinuiy nf lit wjinfi^iinl lohn nuiiili muni ini|i(ii'tiint ** >

wan (lotiiotiid ut an early Htiitf in flio nn'i'-- '

In (tin I'li.'in uf wah-r tlnin in Um nimi of nmnniry,

Uii-rnml uinidiidtlvity (salnriniotisr, (living to tho lar^o inaHHof mi't '* >
'1'lui ivnnoii of Iliiu i'i (InUi tlici
tho Hinall rato of Him' . * * * l h
in tho lluw tulio, tlio
(if wnlur lu-ltij( I0(r in tiinn.i li-*i Hum that of murmiry *" *
tliorolativoly higli thorniiil donduotivity
**f r
thn (HKiili'iitfi nvcrn^lUH of Hi niillliw Ituniicniluro
in liiuru dcpinnli'iit mi HID mi funnily "f Iho Hpinil i liijiiid. It wan prnotidilly oliminati-d hy I* Hi***;
" * ***
1 '

mill Lhn i-innjili>( cliniinatloii of any in- tho ffrtMilor part of tho uiifflow tulin fro.ui
end uf tlm viioiiinn-jairliot with ]nim(Hu v*-ji -*-
In iinliH' lo uliLaiti a porfdnl- fit fov (.lio leaving only a Hinall piiHHa^o for on I 1 1* "** Um
=- -i
of iiie.ritnry. niiuii) tho (iiindiKitimi I*
- willl tlinir Hjiiml HOIDWH it \VL\H niusi'MHtuy

t l** +
([ml ilin Ixii't) of tlio oufllnu' tnhii should 1)0 ms vory Hinall, and nearly iiuUipendenl; f

nniu'ly unifurni n puMNiblo mid iHsniiriilnly


.Id \\i\n must I'HHinitiiil that thoi-ti (1) VACATION

ow TJIK iSrKtui-'H lh:.\ * **-' i

nhnuM IK; tiik iKiimU'ldtiiin ai ilm puitUfi of

jnniil-iiMi K mul I'' with l-hn

vaminui-jiutkM, tlio npenilic heat uf nienmry in teinm nf
1 1 * * w*
* * " *
mid thut Mm nxliH-jiiil poHiniiH nf Iho tnlicH wan eiiiloulated fruni tho oxponini'iifjil I > *

jM'l, ll'l) Hluml<l not 'ID nf (tinulhu hum limn 1

tiikiiij' tho value, of .f
eqniU tu -t-lKill !"* '*
thermal unit at Ifi-fi", wliieli was tho loin " "

jKirUoiiH inniilo tlui viiounm-jimluii, though

tliy j

It would nut nuidor numli if Uioy wuro a littlo tiiro retioiiinmndiid by (irifiitliH at I hi* |'**i-*
('lun^riwH in 1000,
Tlio oxporiniciiliil rennltn littid tu llu* t <

tliuiiry ofUm KliininiLiidii oT tlio lutal; II>HH in Hr-H-J'(l7^x ]()-
M"(KHJHfix j()
- ft
LhoHUsiuly-llow muthoil of calorinmtry IIHMIIIIU'H
if tlio I'loiiI'Md oui'i'iMit. mill tlwi How of .

IID HiinultiinwinHly varied in miuli a

nuiuurr IIH En ktiop Um vim) uf lumnomljnro tlio
HHinc,Ihw liuiifc IOHH Iiy t'lidialiun, dto., will Tliia givim for tho totiiporatnro otiofllott'H t
mmiiiti tioiiHlimt. 'I'liii
fixiiisriniontal remilts any Uinpeintnro t tlio

('iillcnditt' mill BnniOH nlum- limb Llita 4;<>ndiU(in

in very iilitHnly HiitiHllocl In Um mutlind, ninl they

DiUtniiittmL ivmillH of Uio invt!flti(?alin
till l-lui

mi Uiin iinHiHii|)Uini. I'll \vnn mitiiicd, liowovor, and for tho averajjo eliun^o por def-ixsi t\l ;".
Mini- Ihni-'o WIMO Hiiuill tho value - OIKKWH).
nyH(<>!iiaLie divorRoiicou
lit Mio osjioi'iimtiihd vc'riliimlion [or tlio smull Tlio data ubtainod in tho oxjiorinu'iit H i*r-f
HiiininiiriHftd in Tablo L, p. 55.
[loM'ii wliioli, Unmyli iiniiniiil.in unly Lo a fuw
In HI.IKJO, ivmivfti'l (direful examination {fi) Sj'KtJiFro ll'KAi
mi indlcnlionn (if (fiuiHliuil: (^rnn'H, ]']. H. flriffitliH dolor mi nod tlio Hpei'ifl^ t*.*.-'.^t
Ho luiiff ii" Mix dtiitriliiitiixi <if lotnpnratuni uf aniline ovor thoran^o lft to fiO ('.
li.V i*r*^ tm
Um IH the of an appni'atiiH olmilar in its CHsontiiil fintt & g- a -,..
llmmgliuia niijifiratiiH iincinnilitly
name fov Uio nanm ritio of turniu'ratnn', wliat- 1
I'kil. Mail., -Tan, 191)3; ftw, (Jamb. I'Jiit. ,v^ t]
UlO flow, tllO lllt lliHH llUIHl ulu 1)0
JSliC, vlll. iiart !.

TAHLI-; I tho specific heat of oils over a wide rtingo of

AT toniporntiiro the apparatua shown below (Fig. 2)
'J'l-ari'ERATUHKS was employed by tho writer.

Annular Morality
Rotating Contacts
Heating Colls

to thatshown in Figs, a and of the article

"Heat, Mechanical Equivalent of."
For experimental work in ciilorimolry auilino him
various points in ils favour. It Una a low vapour
pressure lit ordiimvy tempera lures, ia a good oluclricnl
insulator, and has a low heat capacity,
On exposure to light it becomes discoloured, but
no in formation is available to show whether this
nffcals the thermal capacity.

For the variation of tho specific heat with

temperature Griffiths obtained tho following
equation :

Si-O'filBO + (*- 20) x -0004 |.((-20) x -000002.
Tho ngreomont between this formula and
tho experimental results will bo seen from tho
table below :
At room teinpoiuturus tho oils were ox-
TAIILK II ceedingly viscous, mid eonmuniontly it was
necessary to employ somewhat unusual
methods of ensuring Uiat tho contents wore
woll mixed. Tho boating noils were arranged
in the form .of two nat paddles HO that

they wo ro in continuous rotation through

tho oil suitably disposed baffles further

assisted tho mixing oE tho contents of tho

calorimeter. It was nceem-iary to lead tho
current in and out of tho calorimeter by
means of two annular troughs of mercury
into which uontaot bars from tho healing
coil dipped.
In tho oourso of this work it was observed (7) 8i.>.K(jii?io HUAT OF LIQUIDS USED FOB
tliat tho volume heat, i.e. tho llBi-'JUttBRATons. Osbom ' lias developed an
specific heat
multiplied by tho density, was practically con- apparatus suitable for the determination of
stunt over tho range of tempera turn investi- specific boats and latent heats of the liquids
commonly employed in
gated as shown by tho [allowing results : refrigeration work,
such as ammonia, C0 a S0 a m ethyl -chloride,
, ,

III and ethyl - chloride. Such determinations

present greater experimental difficulties than
avo mot with in work on liquids at ordinary
pressure, since these materials have a vapour
pressure varying from 1 to 70 atmospheres
at tho temperatures at which tho
properties are of importance in engineering
work. Consequently, in tho design of appa-
ratus for experiments of this character great
attention 1ms to bo paid to details of con-
The Calorimeter.
(0) SwioiFia HKAT OK OILH. In Homo ox- (i.) Briefly the arrange-
porimonts in wliioh it was desired to determine 1
Jltill. .iitti: SMa., 1017, xv. m.
mmit i as follows Tins matoriEil to lw investi-
: iron alioiit 0-3 nun. thick, extending to wilhiii
"",,"*.. , D

gated i isnelosod in a calorimeter with thick 1 mm. of (ho surrounding cylindrical wall. *-

J -, t ,

molnllio walls of known thermal capacity

vanes nro for tho purpose of promoting th\ M'." ' *
(.rihufion of heafc within the annular space eon 1 * * ' .
H'diit in applied (?leotrically and tlio "*
Tho vaiie.H i "*
([''iff, 1
" '
the nmloriul uiulor investigation.
IN maintained tit tho same temperature
jiu'Jaifc *' J "
justabove tho top of the central tuho. At I hi" I
by MIC umial adiiibiiLin firmiigoinont. are two flat circular bailie plute.s, soparal)'<l **
An uir HHIUSO between the polished nickel 2 mm. by tlireo Hinull steel studs. The lower ] > 1 ' * * * '

HurfjusiiH uf eiiloriinotw and junket furnishes uniled to tlio tops of tho radial vnnca with til*-
tlwirniiiL inmihition, Two tiibcn oxtoml from central hole in tho lower plato and several Ii" '"'
fcln> Lu]i u[ {,ho (mlorimotor through tlio jacket tho upper one between centre- and oul.sidi* f* 11
and Him id to tlio outside air, terminating in a tortuous pnsgngo for vapour coming from I M
VOJVOH. OIKS of tlieao tubes is intended for These two plates aro intended to inleiTC|>l"
uoiiiuMiUou to prc&aiiro-iudicating apparatus large drops of liquid which might bo thrown l

and Hie other for tlio introduction anil removal vigorous boiling, should it occur, and also iiml*
thermal shield for tho top of tho caloriuu't*
of tlio nifltorinl to bo investigated.
second wot of four bafilo plates of nplwiel <
Cum \VIIH n!t i>n to nviililhaving heavy metal * " * " * '

separated by about 2 mm. aro attnohed lo thi'

1 h '

oojini;(iliot)H Hpnco mid by suitably c '

nuiort!) l!i<> uii'
fiiii'faco of tho conical part of tho calorlmctfi"

Each has a central hole and four ulntu *-*

* *l
edge so ns to avoid trapping gas or liquid, lmt-
passages are so sized and spaced that the um
III. !'
' ''

through tho plates is very tortuous, BO an to

i* 11
difficult the passage of liquid partielen from t>*'^
in a current of vapour being withdrawn I In
** *,M ' 1
11 "

the outlets in tho top. Tho entire inner mirf!*'"1
-Supporting "'
tho steel shell and of tlio various plates wilJih*
all tinned, using puro hloolt tin.

Method of Experiment. Two Uinlii

(ii.) *
methods of experiment wore employ im I
tho first method tho heat, added to u. tif>
amount of tho auliatnnco uiulor test (imiUr
in the calorimeter under saturation comlit-i
* *'
together with tho resulting ohatiffo ill-
porature, arc measured. By using ilitt t. f "'
tho specific volumes of tho two phason mi*l
latent heat of vaporisation, tho heat !**>*<
tho vaporisation of tho liquid is
Copper and can ho allowed for ; thus tho
heat of tho liquid when kept satunit'*l l p

Ring In tho second method tho onlm'iui(^i*r i'*

kept full of liquid at a constant iirt'Hun


The heat, added to tho variable amount ii t t 1 1*'

calorimeter, and tho resulting clwngii in l*'(*i-
pora turo aro meamircd. A coiTCotion {* t* tlm
heat withdrawn in tho expelled liquid IH * I<?1 '!.'-
JNo. !i. HccLliMJ oE Calorimeter audJaokcl], mined hy special oxpoiiments. By im< *t t lu
data for variation with pressure of tln> Jnt**i[,
dJtitHbiiting Ihoso coiiiinoUoim
which arc nccoasfu-y
heat of tho liquid, obtained from Jiojnuvilo
ovor tho surface, tlio part of tlio
measurements, made with the same uj>}un *it ir>i
llipi'jiinl IwiUiigo tine la lend <jomluolion wns consider-
and material, the corrcotiona for |>rt"i;t.nti \~f
My jmniinluod, Tlioi'Jiiooloinctila indicate rotative

Hiirfaco lompcrniliirca o( tlio jacket onlorimctcr, 10

variation aro applied, and thus a- m*ij< ml
cnoli surfnoo. TJiia determination of tho specific heat *tf
jtuiotloiiB boing illfttributcd ujion
and saturated liquid is obtained.
pormltB of control over the thermal lonltngo
t!io uorreislioii for Hiioh lonltngo as could not
bo Aa a final result, tho apcoifio heat a, In j
uvfjiilod. per gram per degree centigrade, tif
Tlicnnojiinolldiia plnceil upon Hio connecting tulica ammonia, kept saturated, at tho tom]H>v*it.m.t'
imllcftio tlin tomni^mtni-o of llioao tuliM at sovoral is in tho -45 to -Hfi w c\ ly
0, given range
]ioitit <m
HLO jnoltot and in this
\m\ni-n tvliulircs lo rt
tlio equation
\viiy MIC tiiinj)ori(tnni of Uw vuponr oxpollctl during
Vdjini'Inalioii nxjH'i'i'mitnlj)
could bo found.
ILlin Iiifdiln f HID ^cntml tulwin tho oalorimotoi' is ff

iioocNBlhlo at Uni ln)l,Uim for llin introduction of tho

Tlio two curves in l^ig, 4 show tho
hcnUiitf coil and
tfLcrniomiitor. Upon tlio o(*ido
of thin tuho nro fa^toncd 12 mdtiu vanca of tinned graphically.


Si'Eortfio heats of tho motala aluminium, tin, copper,
^T, METHODS. Very little work lias boon cadmium, zinc, load, and silver over tho
on tho determination of tho specific heat range -160 to + 100 0.

lid substancea by tho electrical method TAIII.I; IV

it lor tho metals. The- method, of course,
itself admirably to tho determination of

jjooiflo hcata of good thermal conductors,

vitli poor conductors special devices nuiHt
lopted to ensure uniformity of tompora-
fcliroughout tho material under test.
3) GAEDR. Gacde appoara to Jmvo been
first to measure specific heats in this
lor. In his experiments tho specimens
3d their own calorimeters. Those wore
.tiled to a cylindrical form and a deep
nl core bored out. Into this was thrust
pjier core, wound with a properly in-
3tl heater of oonstantan ribbon and a
anco thermometer of fine copper wire.
:iial contact between the core and the
of tho well was secured by filling tho
ironing space with mercury, using a thin
shell when, necessary to avoid amalgama-
TJiis calorimeter was suspended in a
.lostat and heated through an accurately
tired temperature interval of about lfi a
measured quantity of energy supplied
vy few particulars of the investigation
been published, and tho data obtained
.immarised in Table IV.
LO) METALS. Professor E, II. Griffiths
Dr. Ezor Griffiths studied a tho specific

fi'ljs- Zcitschr., 1002, Iv.

ftil. Trans, Hoy. Soe. A, BOO. 10J3, ccxlii. 110 ;

?oc. Froo. A, 1014, Ixxxlx. GOl ; Phil, Tram. A,

-OH, p. 310.
1 '

(\.) Twiipwaluren IV'. 1 00".- -TJio nppaml.iiH noil boing ndjimlcd iniLil balmnso was nlit i*

Uiu b^utiny uoil WIIH "*"

cmployi'il fin

(lolitniiiimlionfi in thn nuijfo 0" Tlii) ri'siHtuntio of

tit lOll" 0. m nluiwn in Fiff, fi,

Two ittmilivr liliwliN uE thn nintnl under fr.nl. in, ,i t i

M'uri' iniH|)i'iuli'<l in two limns Miiiliwnri'it im- ii.)
Low Tf.m-iifmlurf.it. Tlio oxjtovi .'

low but- ^*
worts isimlinwsd ni toiniujrutui'tiH p ,
a niodifiml form of <ip|mnitiiH slmwn in /'*' '**
an il^ wan very dill'iimlt t,n obttiiii nny "<*' ll ,'!***
" *

1 11 11
hi Una (ippnratiifl
u (sonstant '
wna obltiiniHl liy tins IIHH of si !>>' >

" "'
walled (Hipp

coil of

eoolod all-
lutwl. Tin

1*1 il

iHod in a. * * l *


. A i
^ l! "
* * ' M
f ^ MM *

to DOIH) HJH. r-'t

ii J l IH-H
w[. in.,
* * * *"'
onltiri'd iiil
i'liaii|;<-i" _
A, /**?!'
ThiH inU't'.-1t*tt ( '

lilt wan *' " *|-

alriu'.lod nT * mil-

.Kici. 5.

iti a nmmliuil loin [JowLim* batli. in

llir (icmlral liolu n[ cimli MnoU a mm
t]i ill \vliilnt Lbn tioiixlnl hole" omitiiinud

'J)lio Uiifd hnld WHH iihwl for llio jmi-puHo ot

Mui hltiiik ImliMV Lhi* Hnmninilinn loni-
irwM-Linn of u l-liin-millril tutu*
liy llio
iiij.; ol-lmr innl ()(iniiii(il.inl In n walwr
Mm lilontiH WRH
In i'X|jii'lnicntN oim nf Hio
Uiron^li u nuiMi' fi""'" "i" iJt'Ki *'*' I'l'lnw

ItimUitl 1

1-1 in
lt>]ilin <ruluw
nT Un <iii'liiiirn Lit oim ilfjfivo
iiljiivii llni ImuimniUiHi of Lin' oiiisliiHiim ly a

iiiiiHHumd tmiii'ly >r cli'oiviiuil c-iH'i-^y,

Tltw tf>]ii|n>nvLiii'
Inlrrviil AVIIH iiiuiiHiinHl on
ti ronSMiiuiiKi l'i<lno in Um imiuil mitniior. Kitmo
Mm \t\vit rcMiHlanoo t]Hnnnmotcrrt woro tidjimttnl
t.n oliiKo o(|uulil,v imd iiuulo of llio HiinK) muiuilo
nf wiro, U* ImlaiKiii jinhvli on Llin lii'lilgo \vii-n

\vH ill llio cniili'P of lliowlvo al

nil wlion llio liliidkfi woro in
liMnjicrnhtro ni|iiililiriiini willi Um wmhiHurc.
llt'iino no imsilliiry titill wen* rw|iiiroil in tlvn
M'lJiMifHlinio'H luidKO dirmiit linyniul l-ho c<inu!
iutlo HITIIH,
TJio oiini^y Hii|)iiliml to tlto limiting coils WH miry solid drawn copper tubing In, | ft

Um ooiloil in Llio form of llaL HirU.

inoftaumil ly ImlimuliiK iiotcnLinl difCoronco
oosflivo laj'oi'8 of tlio coil worn Hcpnrettful
lit lift uiitln nKftliiHb tlio
I'l.WVh nf a norloH of .

etripa of oardboard and tlio

ontiro coi\ itn^
oclla hi aodon, Llio ourronl. tlmmtfh tlw

around will* heat-himilaling material. From wollH, and whon il.i

tempera turn liad nearly readied
desired point the euld air circulation around
the i n torts litnigor coils tho air was carried to tlio tin; tlio

valve G, by moans f which nn observer Con-
cjicloHiii't* Blopjicd.
Its woidd then rise rapidly by
tern pern turn
trolled Uvo Hv, excess of air being discharged
con il notion from without and BOOH pass that o{ the
at tho Hafuty-valvo on the eomprcHsnr. After
block which, in consequence of tho slow trull Km fusion
expansion tho
nir circulated through the coil
of licat by radiation and cnnveotiou, would lag
of lend tubing A
and then back over the behind that of tlio walls, Tho lomporaturo of tho
surface oi tho interchange! coils. On the

cnolosiiro walls would then bo niiiintiuncd nlcady at

exterior mititieo of tho thick-walled copper about threo degrees higher tlitin that of the metal
enclosure I'j wii9 wound a layer of insulated block.

copper wiro ',

whUili served as a rests tnnco
Some time had to elapse before tlio conditions wore
thermometer. Variations in tlio temperature Biiflicionlly settled to jimUfy tho ooiiimonconiciit of

of tho walls of tlim enclosure were rendered an experiment.

The lir.-ifc group of readings connisted of observa-
visible by the movements of a galvanometer
UOHH of tlio rule of rise of lcmpcrn.turo of tho blnolc
.By controlling tho flow of air tho

spot. ratlin tion, etc., tlio trannils of tho

by loiniiovatiire
oaeillatiiHiM of tho spot could bo kept within
being observed across HiicoeH.sivo equal intervals
narrow limits and, inidor normal conditions,
(of about ij'jjlh of 11 degree), tho time hotivcon HIIO-
the oscillations did not exceed a hundredth of cesHive transits being of tho order of fiO Koeonda,
a dcgrco in. amplitude, The electrical supply waw then nwilolied on, and,
Tlio intuL'ior of tlio wooden vessel M was after allowing a little time for tho sotting up of a
packed with ulag wool, tho passage for tho steady gradient, transits every fifth of n degree were
withdrawal of tho copper enclosure being kept (niton.

cleat- by a cylindrical tube of cardboard N. Wlion tho temperature hud risen two or throo
degrees auovo Iho HiirrounilingH tint electrical supply
'JL'ho space between tho top of tho onclos WUH switched off and observalioiiH of (onrpcrnturc and
lire and tho outer lid was filled by wrap-
time continued.
ping felt inntting around tho glass tubes
The tcm]):iraturc would then fall steadily under the
influence of radiation, etc., tho rate of cooling buing
Tho blooli of mottil G was suspended within olwerveil in pmoinely tin; same manner us Iho rate
tho enclomiro by a single glass tube H. The of rise of temperature before the electrical supply
eontre liolo uontaincd the heating coil 0, of was switched on.
manganm wire wound on a mica rack and If <j in the rnto of rise or fall due to radiation
immersed in. a light, pnraflin, usually petrol. for 1 C. dilToronco in toinjioratnro hetweon tho
Tho boating coil -was fixed to a short taper block and tho Hurrouiulinga, then assuming
ping of copper K, wliich cloned tho eontrat Kowton's law to he valid for tho loss or gain by
hole. '.I'bo ruHintanoo of tho coil was about 20 " "
radiation (an assumption which was fully
olmiH, A platinum thermometer was inserted
justified by tho experimental results), wo have
in tho cylindrical hole T, tho annular gap
tlio oxprefisicm
between tho stem and tho walls being closed by
a packing of nshi'sloa thread. Tlio differential
arrangement employed in tho previous experi-
ments WHS abandoned ns it would have required for tho rate of rise or fall under tlio influence
too long n time to obtain tlio equilibrium of
" "
radiation nlone. Hence, plotting dOjt)t
conditions. against 0, tho straight lino joining the two
(iii.) Method nf flayer
iment. In these experi- groups will out the temperature axis at =
ments thu practice- was to heat tlio material which determines tho temperature of tho
through u Hinall temperature interval from aurroundingti.
below the mi rroun dingo to an approximately .for the rate of rise under tho combined
equal iiitorvtil aboyo f and observe tho rate effect of tho eleetrieal supply and radiation wo
of chiving this period.
rise Tho method of have tho equation
experiment; was such that a direct deter-
mination of tlio temperature of tho enclosure
was not required. An experiment was con-
ducted o.s follows :
where E a/Il is tho electrical supply per Rooond

Tim tcmpemtm'o of tho enclosure was lowered

in thermal units, MS
tho thermal eapaeity
from tho of tho bloclc including that of tho resistance
progressively by utiliHing Hie full supply
coil, etc,
compressor imil controlling tho How so as to produce
a steady prcanuro drop through Uio valve of 120 to Plotting tlio observed rates of rise on the
" "
IfiO aimcKnihei'CH. Btinio scale as tho radiation observations, it

tumpcmture. of Uic block would fall at a steady

Tlio isobvious that tho straight lino thus obtained
rate by radiation nd convection to the enclosure should bo parallel to tho lino joining tho two
" "
It la prohnblo that wool In Ua nntnrnl stato would
groups of radiation observations, since tho
liavo lifion better Itwnlnlor at those low lenrnera-
tangent of tho angle made with tho axis is
tiircB, shuwi ilin fireaHO In tho wool prevents It from " "
absorbing moisture. equal to tr, For = tlio radiation term
vtmmlu'N, honco, if D(?N/(")(. donotos tho value of TAHLE V cotttinwtl
the ordinato at this ])<>int, then
< 3fl
' .K~ W llcafc. I H
( 't '

tJi iuiw
from which & can ho oblnincd. Tho results fioniiiM (MOLTEN STATE)
obtained avo MiimmaiiHcd in Tablo V, an<l shown 0-3234 7-H
in /''(';/. 7, whoro T is the absolute 4-1 Ml 0-31HU
300-1 0-3217 7 !()

(1--2 g (11) NEUNST AND

Q-O ohsorvers made a Horiea of point to point il< ' '--** 1
minations at very low tonipornturoH, imii**' Jt
calorimeter developed by Euekon. 3 " A ''

metal of suitable ai/.o was shaped into u lie *i ' * * w
0'4 t-' 11 *
cylinder and a loosely fitting core miulei fui'
y o-a same. On the core was wrapped a plulii"*"

-Co-o TYio ialtar H donotea tb& absei nations wire, properly insulated, to serve as a ruHmtJi-t
" ''''

of Hamst In ilia onto of allwr mul of

04-0 la a it.
thermomotor and also as olootrio lioator,
core was placed in tlio cylinder and
0Ceaf. WOOent. poured into the crevices to improve tho
Absolute ',
Temparnliiro \
contaot (see Pig. Tho whole was Himptit i * I*
100 1UO" 2SO U GQO 300 QflD" 3QO in vnctio, and tho speoilio heat ovor i-*i'"
l''IU. 7. tomporatiiro intervals
detorminod from mensnro-
loinjjoraturo and ('! is tho atomic lioat, i.e. inonts of energy supplied
heal- miiltipliod by the atomic weight :

electrically and of tho

tomporatura rise roflulling
therefrom. Nornut ami
Lindomann applied the
same method to poor
heat conductors. For
such materials the design
of calorimeter is shown in

Jfig. 9. Tlio wire was

wound on a silver lube
projecting into a silver
vessel, tho high conduc-
tivity of tho silver assist-
ing tho equalisation of [,-!, H.
tho temperature through
the mass. Some of tho data ;

rnotals obtained by Noimt aro

(12) OoH'iiii, Harper
stndiod U
specific lioafc of copper over tho range
16 to 50 t!. Tho specimen was in.
tho form of copper wire, which a]m>
served ns its own tiiormoinotor and
lioator. Tho wire was fiO metres in
length and- 2-5 mm. in diameter it ;

was compactly coiled into n number

of Hat spirals separated hy mion,

plates. Tho ooil was suspended in

vacua, and heated with a measured ,.

quantity of energy supplied elec-

trically, tho rcanlthig temperature HHO l
measured hy tho change of resistance. '

Jotirn. tie PhuHiqw. 1010, [-1], Ix. ; Silguuff
1010, I. 247, 202 ;
Ann. </. 7*fty*., I
PhusA; Zeitschr., 1000, x, (J8fl,
Sci, Paper Hitr. SMB., 1014, No, 231.

results of 27 determinations butwcon Ifi and This eocllioicnl is l!u? menu of Ihcisn u'nen liy
fiO uro represented by tho equation recent olwoi'viT.s employing isloelrio luiiliug jnul a

S = 0'0917 + 0-0000!8(f- 25) calorie^ ]r jioinl lo [mint mettiml <IH shown in Tiiblc Vll.

grain degree. TABLB VIC

4-182 joules is taken as n<[iml lo <mu 20 calorie.
(ii.) Comparison a/ Data !>// Varioun Ob- Observer.
Cnlorlra pin

(li'iijn I Ji'
36i- vers. liar per lias tubulated tho data given

by various observers for tbo HpceHIo bout of 0-<10fK>H)

copper. In order to compare the results at one K Jl. Orimtlin ond\ O-OOOO-H
definitetemperature tho coefficient (M)00(M4 K/er (JriDHlis j
has been used in reducing results obtained at Harper .... 0-000(|.|K

It will bn CI|).MOI VIM] from

NEKHST'S VALUES AT Low TEMPHIUTUREH a ooinpiirifKHi of (lio datit

gtvon in '1'ablo VI11. (p,
42) nnil wlimvn griiiiliuinlly
iu Fiff, W
tlmt Llioro in

tho rc'siills iE
observers iialng tho oleu-
trioal jnoilind both as ro-

gards tlio obsohilo valuu

of the H|icoifio Iioat and itn

toinpovatiiirt! coofliiiioiit. It
very iinprobubHo Unit
thoro in any ByHUstnutio
ort'or common to nil HIIICKV
the tliroo inotluuls clifim 1

nidieally iu (Jdtnil.
. T
!', .

liBiires litivo iJOfin obttiineil

!. "-."'"I
from th
" <)C Mirilinl L
ima Urn aliovo
tsurvcs given In tlio pnpcr.
t ],

Hpeoifio heal of at
tho various lorapomturea to tho fiO 0. value, diftoront tempera-hires 1ms IHJOII dovolupwl by
tlio foi'jinita Callondar and liia nssooiatL>a.
S = So i- 0-000044*
Gases present grontor practical tlLfliouUios
100 ito" iarf tad

O'OOQ -5

001 =-

o-ooo -s-

-20 -10 o3 10 za> ;io 10 DO" DO" 70 ao 00 100 no'

Fro. 10. R]ii!Dino Ilpivfc of Copper.

boing nasumod us valid for valuer ()I ( from than either solids or liquids, since it in nocrsHiuy
to 100. to tdko grout prcoaiiticMia tn oimuro
Set. 1'ajier Jiur, Stits., 1014, No. 231. of tomjiorature in tho gim
'L'.uir.K VIII caulinued

hydrogen ncnlo.
20" cnloi'Io,
Coppnr 09 -ill JUT font puto.
<>\]triwa<Ml in fnviiuiliv

liivio (jqunl Hi 4-1 SB jmilcH.

\Comiii( roial
drawn copjipr. Caloric

J oiiiuil t(i 'J-lfifi jmilcH.

Ciiloi-io prpial to 4-188 joules.

(Assuming ico-poiiit=273'lIC.)
Elt'olroly tio copper. In IcniiH of n
onlorie equal to -J.-18S joulca.

Very pure clend-olyl io ooppcr. .ISIeotrlo

hcfiliiig, point U> point method,
HuHulls ivxprcNSisd l>y fonnnla
c-0-flt)l)fifi (l-t-O-OOOfiS'iU- 0-CVIHJ )

Kcsiiills ('xpri'Kscd ly formula

Thu unlta In terms of which tlui ahovu results avo oxnressiiHl uro not all (U sumi!, hut tliti illlforaii'iM nncil HOI
ho tnliea Into luiooiint In maklns eomimvisoim. In overj' oaso the illfforonuu between the unit employed anil
the 15" euloiio or tho 30 eulorlo (whloli illlYor from each othor by about one, imrl In * Ihouanml) In lens Mian (ho

Otillomliir and Swnim 1

ap])Iii!(l tho oontiini-
cooled. Tliia i.ulio was packed with tightly
ous (low inotliotl to tho dctoriniimlioii o tho Jitting discs of <n>ppor gan/,o. Tho gas under
lionts of air and onrbim dioxitlo at test entered at m, was heated up to tlio required
atmospiioro pressure at 20 0.
A temperature and entered tlio spaoo round llio

and 100 0.


(i.) f'lie Ap]xtmls. In those

oxporimonla a steady stroiun
o gas was passed through a
jaokoted tube (tho eiilmimctor
proper), in wliich it was heated
by a current o( elcotrieity paa-
ing through a platinum coil of
1 ohm resistance, tho rise in
temperature being measured by
two 12-oiim platinum tliermo-
motors used differentially. nllow exi'csH nf
|ian to CKCIII.WJ lilt,
Tho calorimeter is repre- contnlnlnR milli! K01E anil <'a(!l a ; l!, cutliHi wool ilniiL

sented in
B "
filter; I), luitoinatln jirwmiro KUiilaUii 1
; I 1

',, throttld:
ditigrainmatioftlly 11 [<1
' ri F, lunci mckeil wllli BOIIKO to lirhw HUH to tlia UeHlruil
/''((/, 11, tho hooting ooil and '

l,n i n lie rn tuns ; <), fluo molnl tuln'S for mcnKtiroinotit of

platinum thormomotors being situated in tho How; 51, ollKaiif(,

tube AB, which in jaokoted by Ibo tube it, Thu

cnlnrinictor propoi1 . It next puBaotl through
tube FGt formed the heator iiv whioh tho gna
the tubo ?i, into tho ealurimotor, and finally
attained tlio desired tompoiuLuro, and boing
double-walled could bo steam-boated or water- emerged by tho tube p.
Tho general ixiTftugonient nf tlio
Phil, Trans. A, 10 10, uex. HH), will bo undo] 'stood from Fig, 12.

(ii,) Theory itf

MctM.W (! is t!io clrctmi current. ire greater hy about 2 per cent than
.K tlio potential (lilformu'Q between Iho rauln of the irrcsponding values found by Regmiul *>
heating coil, SW (ho rinn in temp Mature of the gas, Q by Inter investigators who have cm
pi* *?*'*"*
-' 4<'*
I ho rnto of llmv of the gas in grammes per second, Methods similar in principle to that of

tho mnclinnitsiil equivalent. nf heal, nnil 8 the specific juanlt, but it lias now been established t' 1Jl ^
heat of the gns at cons tout piTRanro, tho elementary 1' J% *
Llcgnault's method gives values which ni'O
theory of tlio espnriment gives
jy about this amount.
(14) Si'Kcnao HEAT OP STEAM, 'Hi"
ivhoro ft30 IN a term representing tlio heat loss by worth 1 developed tho same method titV
radiation, etc.
determination of tho specific heat of H^-
A similar experiment with a rato of How about at atmospheric pressure between 10<l- (*~
bait the nbovo vnliin, and witli Uio clcolrio onrroiib 116 a
in tcmiioralui'o was about,
adjusted fio that tho rise i.) Outline of the Method. Steam is fjjo
the name a before, Rft-vo n, HOOOIM! equation,
a ted in a boiler and thence led to ono liml >
that ft could bo eliminated and S determined. J-tubo pressure regulator. Tho prosmii-'*
Tho torgmt ourrcnlfl <>C 8 through tlio apimvatiw l *
tliosteam forces tho mercury down in thin 1~ i i >
were of IJin order of O-fi litrti ]icr fteconil. The ralo of
coiiBlnnt by nn aiilcmiii.tio pressuro
f Iho U-tubo and up in tho other liin
flow was Icepli
which the adjustment of
regulator. It was i)irauvetl by pnasing tho giis
through 1(1 iitu> metal Lubes arranged in parallel, supply of gftH to tho largo
between thoir burner, used for boating I*"
niiil observing tlio pnsisiiro diflWcnce
ciKlfl, Uio menu pre-fiaim:,
mid Uio Innpm-atnre. Tim water in tho boiler, is
riilo of flow in terms of theso After passing tho regulati *
expression Rivinfi Uio
was found by a wiries of ojqtorimcnla in s leant, now inainlainiid
into a reservoir of about;
wliieli tlio gaa was pumped
50 litres onpaoily, and llioii discharge allowed to
through HID apparatus, Hy means of a special
of Ran
device, tho limes Inkcn for certain qnantitica
lo pass through tho apparatus wero recorded auto-
HO that
nmtically while tho gas wan notiially flowing,
tho initial fhictiuvtioiiFi ivcro avoided.
Tlio value of tho clcotrio ourreiit was obtained by
set up at between
measuring tho potential difference
tho ends of a, standard resistance c3l In lornm of N Fro. 13,
cadmium colls. Tho heating oftoots uf tho Iradn of "II

Uio healing coil wcro dotcriiiincd by oxperimontH rfa, and rf 3l (lmliwforc.omlonae(l \v4itor.
mado under tho oxaut condition!! of tlio main experi-
ments. constant pressure, IB led between tho
Tho tcnipcmtnro in the main experiments
lino In tho jaokot surrounding tho calorimotor i>i.-
wan about 6" 0., anil it was measured to 0'001(J. thonco through a separator and a throttlo
Thufl tho apooiflo hea1 ii'oro meimurecl practically
the spaoo enclosed hy the double-waited j 1 1 c It
I , t

at Binglo temiiomliires instead of over lurgo ranges. whonco it passes down tho calorimotor llmv-
of aflminimR Uio licafc Ims for a given
TJio validity
lomporalnro lo ba indopondciil.
in (if the rate of
tubo to a oondonsor, Dnring tlio *-f pnw%Ko
tho steam tlirough tliollow-t-ubo ititt ]u;itt*#l 1 *v

flow of tho gns wan tested by experiment. Tho

matter was also oxmnincd from it thenrotioal Btand-
moans of an olcotrio ourronfc passing Uirtvi|j;li

and oorrcotioim worn calonliiteif mid applied jilatiiium lioating coil, and its tempera tin-** in
ivlierotho OBflnmptioiiB made in tlio elementary measured on a platinum resistance thoi-fn
theory ivcro snoh as lo lead
to error* of moro tliaii motor. Another tomporatni'o moiianvonntiil
about* ono in 10,000. Tlio corrections wero made when tho supply of electrical onorjufx ii
snmll, only amounting to ono or two parlfl in 1000. out off, and tho between llui'Mti L\

tomporaturos gives tho rise in tomporiitiiv**

Full dotallfl ol vftnoua other prconutiona arc
tlio moan ol a tho steam,
given in tho original paper, and
largo miinboi. ol observations gnvo
tlio follow- (ii.) Galorimetna Arrangements (Fi(f, lit).
.illrf i
Tho calorimeter proper consisted of a gl IVHH, t 1 1
Air Y, about CO cm. long, in which thohoiUitijj? '*
21173 oal. per gram degree i> 20 0. C and tho thermoinotor wore llxotl. N
100 0, tube was jacketed hy another glass UiVn;*
which enclosed the length occupied
lioating coil and thermometer.
0'20202 oal. per gram dogrco al '
20 0. calorimeter (low-tube and its
100 0.
ing glasa sheath were carried on a split lj|t
0-22121 t

10 Bovoml dotorniinntiona- agrco in. oaoli cork wound with omega tapo and fixed,
to about 1-6 pailfl per 1000, and tlio mean a ateam-tight joint,' into a space onoloBOcl
doublo-wallcd brass jacket. Tlie lowot-
It&iitrQ probably coruoot to ono part in 1000,

values of tho specific heats obtainot

Phil. Trans. A, 1015, ccxv. 383,

of this jacket cowiminientra with a dcmblc- justing Ji in otieh ease so that d<) remains
wallcd side tubo and tlio stonm entering at K the sitme, ft and k can bo eliminated and S
passes between tho jacket walls to V,
which measured.
comninnicatos with tho stonm separator U. Tho value obtained for tho specific heat WUH
It then enters tho inner portion of tho side 0485(> cal.per gram degree at 104-5 C'. and 7ftO
tubo through tlio throttle T. Thin tnbu is mm. pressure; then, assuming ^ t
a linear variation with tempera-
tightly packed with gau/,o dines; tlioneo tho
fitonm passes up tho tnlio S into tho top of tho tui'o as experimentally dctor-

flow-tube at I', and descending past tho heating mitied, this coiTe.spmids to a
coilC! and Iho thermometer flows away into N value 04S78 at 100, Loth ex-
the condenser at ij. pressed in terms of tho calorie
On its passage through tho tightly packed at 2(T C.

gauze discs in the side- heating tubo tho steam

was heated up to the temperature- of tho jacket. (iii.) KJfcet of Impurities in the

Steam. Steam in tlio immediate

In Fig. Ill, whiiili represents dii a grammatically tlio neighbourhood of tlio saturation i\ \
iimuigomenl thus desoribnd, it will bo noticed thai point is liable to curry small particles
tho cylindrical space inside tlio double walla of tho of water in suspension, which can-
main jiioltot m
divided into two compartment*) by a not bo ovnponitecl completely by a
diflo K.Thin disc of brass wns soldered to tho moderate degree of aujiorhcRb if I

inner jacket tubo about G em. from the upper end. such as salt in
any impurities,
Its function in to prevent tlio steam in tlio tubo S solution, arc present. Since 1 nig.
from impinging on tho rubber cork closing tho upper of water requires more than half
end of tho tube, and thus being cooled. a cnlorio to ovaporato it, and tlio
Any slight cooling duo to tho steam striking the heat required to raiao the
lower split cork in of no importance, winco tlio steam perature of I gram of uleam 10 U.
would lai wanned again during i( passage up between is only f> calories, it in necessary

tho How tuho and Iho Hummnditig jacket. Tho whole that tho initial filcani whouhl not
of the jacket, tlio separator, and tlio connecting contain more than I in 100,000 of
tubes wore heavily lagged with felt. A novel feature water if tlio speoilio heat is to lie
" "
of tho apparatus in tlio spiral method of mixing tho found correct to 1 in 1000 over a
ntoam, in which UNO in inailo of A. number of circular range of 10 U.
dines punched to lit (ho thermometer tube and then The ri.4o of tho boiling-point
out along a diameter, bent, and soldered together to produced by a; {iram-moleouhfl of
form a continuous Hpini I Halt per gram of water in approxi-
round tho thermometer mately 1000.B 0. The proportion
(BCO 1% 14). TliiH of suspended water remaining n-
method of mixing i
evaporated at any degree of super-
found to bo a great heat 0' will bo 1000.1-/0'. The
improvement on tho quantity evaporated in heating
gnu xo method previ- tlio steam from 0' to 0" will bo

ously employed. 1000^(0" -Q')!0'Q", Tliis will jro-

Tho usual equation llnco an apparent increase of

for tho continuous tlio moan specific licat of the

steam over Uio range 0" 0' equi-
How method is
wboro L ];m. 1C.
valent to lOOOLff/0'0",
Is tlio Intent beat of
where EC is tho elec- tion. It was found that this extremely simple and
trical convenient reduction formula titled tlio wsulla
energy sup-
S tho required obtained over dift'croiit ranges of temperature with
heat if ex- extraordinary precision, and vecon oiled iippnrenl
iliHorepancica wliich bad previously been attributed
pressed in joules per to omits of observation.
gram C., Q tlio rato
of flow of tho steam, (15) DKTEUMINATION or TUB
dO tho rise of tem- HEAT OF Am ANI> OTHKH GAHKS AT BOOM
|,' K]< !_!_ stoam, and
lulQ is a OUS FLOW ErjUCTitioAij MKTiion. Hohcol ami
term representing the House 1 3iavo determined tlio sjicoifio hnat ol
heat losa. If this loss is independent of tlio air and other gases at +20", -7S,nnd 18;i

flow, then a linear relationship exists between by tlio continuous How inoLiiod. Tho air was.
tho values of EO/Qdtf and 1/Q. This is directed in a steady stream through n pipe in
found not to ho strictly the ca-so, and another which it received a known tunount of heat by
term depending on the flow is inserted means oE a lioating coil.
J'iff. Ifi shmva tho glass calorimeter
in tho
in tho fundamental equation which thereby
form in which School and Jienso used it. Tho
of three, rates of flow, ad- 1
..IBB. A, Phiis., 11)13, xxxvll. 71).

HUH, wliinh in brought to a Htcady timi|ii!ratnro, iii f>nlcr l<> (li.tli'iliulo tlio licut
('(|iiiilly llm iv ' '

onliTH llio nilnrimnli'i' from Inflow and flown uro liinind tdgidlicr and wound ruinx] thn linn ru| '

through a .spiral, then through two ^]HHH jiuskots Haii/n <

HH far tm HJIJITI! (icviuitii. '\'\ui wim miln
J *

(! itml I!, nnd linally nauilios l.ln> ininu


tube A
wlmih {lontniiiH tlio "" '* j
(soil, Tlio (omiioralun ? 1

tint in- and (wUlwviiig #i '

ivory Htri|> M,,. 'I'l vii'c.-i wliinli condiml, Hui mil J
am dulorminoil liy tlm
/j Icil ii|i through Ihn inner
IHIVO Ki, with (ho ivinw (

mid I'n. Tins wholo in Hiir- MHiri'u of OKI (iiirn-nl. and to (lie voltmeter,
M I'lHindiul liy n viiomini and lii'iiting (roil and a|i|iara(.nn f<u-
mixing nni ciml
a (,'liiHd jiir-kot mlvtii'dd inmcto in n In'iiBH pijMi j\l
wliiuli lilniiilo n Hcouncl

ftinl is i!(iiitiiiii(!il in a liiifh [lijiii

Mm "i[
a in ftmtuiu'il with Hi-alinj;-wax iir
lit M mutant toni]wi'atimi.
uninilar j^roovo 1111 till) iniiiu' ]iipo A,
Q, AH Mm nil When nb room doinpni'atnn! J * *

ut lirht flown working

through lli jmslcHtt (! and e.a.loi'ininidr was pliuuid in a lin-go wain-
'A I! Imfoni it rcimln-H tlm mil w] I At low tomjioraturo, on (hn
Hfirroil. i

oalfii-iniolor, it ulworliH dim hand, it was jilncod in a vacnuin vivmd v
of flio hoiil (iotittiitiod a niixtuni of ('(} 2 HIIOW and alin '*
gri'ulor part
tfiwm oil' (.o MID in i ior viusuinti or lii[iiid oxygon, fn wioh UHO dim giiH, lu-f* ** *'
J iKsinmlinjf to tint iiiinciplo (if (Ditiiring dim {inloi'iiiiDtwr, jinwuid through n jii J **
tiwmtor (iiii'i'dtil. mill tliim which was lionlnimid in tlui muno Initli an I I**

(iBHisla ihn iiiHiilniinn (uition

If of HIM jiuiltol:, Thia firmilly Tho HH mo InveHti^al.!)!'!!,
wliint Hhitt.yi *)--'
ll'ltl, 1ft. iwlu<!{! tliii IOHH of ]HHL(., liolium and funmi other raro gustw lo low In** "

Imt <1nos nod ontlroly i>ni- lioraiiircH, nuidilifsil (.]i

ajinaratim n n.-i ^ *

vunt it, '.I'lm lnni|iomtim> of llio oiitll employ a climoil isiraiiit.

Kns in iiHwmmid in tlio tmiiHvo.i'HO Hinvl.Lon Tim ri'Hiilts thoy oM.airiod aro miminiirim '* '

Tho lidiLtiiif? noil in B)IOWII in Fiy, 111 and in T.ahlo IX. T.liny also oalonlatiid out, I I**

nutiHinU uC iioiiHlanLiin wiro K and iw woniid in <!onu4])<mdiii(f vuluca for (lio idtsal $nn ulul*-
' * 1

two HiHtllimm on 11 gliiHH )i)it, 1

Ann., 1. WWH., l!)l!i, xl. -17it.


'jftlmilat(l viihiiw, k and ka (lie raLion fp/du and (-Vo/f-W ninnoo lively.

these results arc inserted for comparison along- Ttui oil cn lio Iicntcd wlicn desii--oil liy
side tho others. mcaiiH of llm roHin(iiuco coil rj u , d in a Hfcntn jiiuhot
Tins values at constant voliimo aro obtained fliu'roniuliiiff llic; ciddrimolcr mid (ho exit \H\W,
tho cntraimn ]>i|io quid) free.
experimental numbers at constant
from tlio InnviriK
w losn u WT
by means of
rirossiH'o l,ho expressions dedmind ]H[iu protoott'd iigniiml. liy

by a combination of
tlic ordinary thermo-
dynamioal equations
ivitli .1). Jiorthelofc'fj

equation of stale.
Tiii'! .Si'iwiMU HEAT oi?

OVI'lH TirM JiANUll 1 TO
1200 ATMOfil'lIHIlKH,
In 1914 Ho! born and
Jacob l
made a now
series of measurements
of tlui specific hoat of
air high pressures by an
electrical inothod, Tlio
calorimeter employed is
shown in li'iij. 17, Tho
castings used in tins
construction worn intido
of nickel Hleul of high
lonsilo strength. Tlio
air enters tho calori-
meter at tho bottom
through asmall spherical
piece c.^
and leaves at
tho top through n similar
arrangement <: s Ho- .

twcon them) two points

thoro a nickel steel

pipe s with Homieironlar.


ends c a and c into (|1

which tho wido pipes c a

and c 7 load for tho
entrance and oxit of tho
ouiTcnfc of air. Con-
nections hotwoon tlio
various pipes aro made
by llaii^cH fitted with

A heating coil ql
through which tho air
flows is enclosed within
the pipe r, 5 ; from hero
tho air passed through
tho annular spaces ^ and
1%, ils direction being
Q Dulfit*
changed twice before it
leaves tho calorimeter,
The spaces /[ and. /,,

avo oneloHod within thveo nieltol walls. Tho n[ (.wiiilod nilk and fruin t.hn aoiion of Llm in
direction of tho air current is indicated by tho iiir
by n gliiHH pijm in it. Tho hwitiT '/i
arrows in tho figure. (ii)iiHiH(Jiof it gmii]) of (10 split niolfol tnbcs ttbnub
-I nun. ionor iHinnrtor and] d-fi onlur diiiiimtpr, Tbt'o
Tho outor Hpauo of tlin (wlurimolni in iilmi divided 1

jirn fiictcnwl to two jinroclnlii jiliLlcH nnd urn licld

into two cylinder* o t and " 2 (Imin^li whioh nil friim

tn^ciUn'j', tini! licliliid tlw olhi'i', liy olipti. 'I'lui nil'

tho Hinall turbino r in driven in tlin dimition of llm
|mti.'u'ti Ilii-Dii^h nnd around Urn liilx'.i; (invn-nL in

Zeitsdir. Vereines Deiilxth. Ing., JOLJ, Ivlll. H2I), nnp|i!i(.'il
to thiH liciiU'i'
by M'ii'i'd intndiiti'd uilb (jhiHi)

Hie Ilio black mol.rio Uioniminotoi'H blM'll mi hlK-'

Ill purl vi-jiroHciiU
that tho ntl
iiii-iru unll (ImHllHillid plll'td illlilllltlmg
llllllt'l'illl. dovfilopixl
'.I'ho a lompimilnro of

ro]irtWH*ntod by tho e-mpivioal


, and tho lochniqno of

tlm jaekot Ima IIMOM gi'ojttl
if fn*" il11
tated by tlio uso of elootrkml heating, tt>'*
' "*
the iiiliorout diflimillira iiHunomtotl
whom 71 in Urn premium in atmdiqilwi'os. mol'hod ,of (lalorimotry now ntuler cdiiHii ->'" h
Timvalim 0'24i;{ obtained fur Mm H|iom(i<! Mini iiuiro than comilorbalamio the lulvnn)
hout at fill" and mm atimwplmm pi'CHHiiro it ofTerw: iioiiHoqiioiitly, it in visry litflo
a^niei vory mvtLHfantoi'ily wilh UIOHO
at tho present day.
liy Kwim ('-ail a lit 20")
nnd School and Jlmmo 'I'ho .ItmiHen iiio ealorimetcr and tlm
(SMO!)). .11. nf (!diii-Hi> dilVoi'H from liogiuuilL's stciini lialorimolur are olniwiisiil
valiH!, wliiisli in now known In ho low. thin motbod of calorimotry.
Tim following tublo giviiH tho valuta oHivinod J)owar luia applied tho mimo
for tlm Jipmliu limit air n tomporatimi of f>i) (.',
to boat determination a at
ami fin viu'inim procures measured
in kilo-
low tonipeiufcurcH, and obtained dutu
gram por Kfj, oin. corning tho mean Hpooilio heat tif ma(<

X betwcon liquid liydrogcii

liquid nitrogoii tomponituri
(2) BlfNWUN'H 1(113 (1,\J
MK'I'BR. Til tllJH

hoatgivonoiitby a in <*
ing from Homo higher tciu|H"
turo to 0" (!, in (il)tiiim'il
oliHorving tlio ooiitniol.inn u'li
fciikoa pla(!0 in tlm (ilmngc fi
to wa(:or produiidd |y

boat given by tho hncly,

Tho ohmtrviid voluino ulu" i
iH eonvortod into culorirn
])iiru<w(!n tho rofliiltH nf vuhio for tho ooiiHlaiit uf
and tlui vuliioH oalitnhitdi! from tho ieo oaloi'i motor, i.e. tho JIHIHH nf rnoroiiry t\\n.

oxiiuriinontH of .ionli: and '.rimniHon, into tho iiiHtrumonL liy tho additiim ii.f

(11)11), and Niiull (IHIH) Hit) liwwrtod. mean ualorio of beat.

ivoro olitainod rroiu Lindo'H formula Numerous deLorminatioiiH of tliin mmsil.;
havo boon mado and tho valuen uro Hiunntit r t
in Tablo T.

wlmro i Mm Hp<ntili(i
lus/it iifc
pi'PHHiirn (I and
S Uin noolitig when tho ]ir(!i!nun'
in roibmcd liy
from p to u vaniHliingly Hinall vulno.

Tlllfl OJIANOH OK H'l'A'.L'K

(I) Tins Mu nioi>."-

ln thin claHH of oidoti-

molriu anyllanocH tho ijuauLlty of liotit to Ifu

moamu'wl IH dotonmiwd in tdrniH of any no
of tlio following :
(i.J '.Wio mnss of ion molted}
(ii.) iitoain (jondoiiHtid i (ill.) liquid hydnigon
or nxygon vnpoi'iHful.
Rui'ili incnHiiromoiils do not rcqniro an
fioourato inciiHiiroinonli of Binull tomnoratiiro
might bo rumnrkcil (Jmi iniiiiy of (hi

oJuingos of tlm ouloriiiiotrio fliiiil, find taking avo bused on uhsorvallimH of Ihi*
111(0 odiiHidoi-atidii Mm Hlulo of thormomotry lo tlio onhirhnotcr by a. mnnll <|ii
half ago whwi MIIH nuithod waa
water cuntninnd in im mvohipci whom
iiciitiiry of ll
inh-iHliHiiid, thin faitt wan iin(|iu^tidiiably
capiMJily wan omnpnrablo with that of llui
ixnil Mu* of
iidvnntngo. 1'iirfcliw, Uiniiinratiiri) water. iSinooin (lie majority ofcnwH im nlU'iiijn li.
tlui Hurroitiiding atnioHjilioni (iiin
havo but IJPOH mailo In vnry tho ooiulilioiiH find tluix I '
t *.-
ofloot tlm indinatiotiH of MID tlin vulurit avo not I'lititln^*!
lltllu upon ByMlcinatiti orrors, all
fliilm'iiiHitor HJ mid tlui initial and final tompom- (Ito fianin weight. For exnmpli , (ho HKHVU Miv*-n 1

ni'O tho flivmo, tn rouont yourn (sidnri- tlio menu of two ox(ii'imonU <""I^tt *

muter precisely similar conditions 1,1 10

glass mvoloni;

(ii.) Precautions in Vac. In tho original

weighed 0-2 gin., cmitaincd (Kl gm. nf water, and a manner of tlie the
performing experiment
platinum sinker (weight O-fi gni.) wan also altaohed.
Dictorioi varied tho
instrument, abnvo, wns plnccd
us doHoribed
quantity <>f wntor from 0-6 in a vessel contnining pounded ioo or siimr, the
gin. to a gm., and eunsecMioiilly liis delcrrmnnUtm
is entitled to greater weight flian dm others. top of tlin tulm A and tlie tulxi S alimo pntjuct-
Griftiths vnluc was obtained ing above tlie ice. Jt IH, however, genomlly
by supplying a known
quantity of heal measured an clootrieal cnnrgy. Tlio
found thut there is always small (JilVorontio
heat mm mijiplied by a mniifjanin noil wound on a
in the fi-flo/ing-point of tlie iuo in the inalrii-
mien molt wliioh fitted tho interior tuhij of flio merit and that of the ke outHidc. If tlie
calorimeter (HOC Fig. 2), and tho rcmiKs are fwi.ied temperature of tho outside it to is higher, then
on tile olcotricnl units o[ KJ[.F. imd resistance. Tlio thero will bo a slow molting of tho ico in the
conditions wero varied. Tliun tho vatn of which will oauso a iscmUimom
energy instruinont,
Biipply in the fas test experiments wns moro tliun
creep of tlio nicrciiiry monistniH Itnviinls tlio
seven times that in Iho slow eat, ami tho
probable instniineiit. If tho freezing-point of tbo
emir liy the method of leiiat s^unrea wan Iran than ice outside ia lower tlian that of Hie ico in
(H per oont.
tho instrument, then tboro will bo a Blow
Bimson employed calorimeter to
his ico frco/,ing of tlio water, causing tbo inenifinus
determine tho latent of fusion of ico as follows : to creep away from tho bulb. Tliis orcop
Known weights of water at a, known boiling generally ^_,_ h T .. ,_,,,. JT
.,-.,.: ,_
temperature wore introduced into tlio inner amounts to
tube of the oalorimotor and tlio contractions li or eontiinetrL'S
observed. hour, nnd is. suflifjient to
In a separate experiment a known weight tnako it very clillicnlt
of ice at 0" C. waa contained in a to obtain
bull), tlio trustworthy
rest of the space being filled with
mercury. iiiensiiromentM. A slight
The ico was molted to water, tlio
temperature ndditiun to tbo iimtru-
being maintained at M'oroiiry WHH drawn
I',. nient will a Inn ml- rlinim-
into tho bull) to occupy tlio sjmoo loft by tho alo tbo (inieii, n.'diiising
ico in melting, and from the additional it to about a tenth ilo
of moroury tlio oimtmotion wa.s obtiiinod. normal value. Thia ud-
Ho found that the malting of 1 gram of ico didion, Hiiggi'Hloil by Hoys,
oauscd <i contraction of (HJ007 o.is. in pluditij,' (ho
I'rom tlio results of tho two Hots of oxfiori- in Jin oinply
monta lie calculated tlio latent heat of water vi'Hfael, tho top of wbiiih
to bo 8()'()20 oalin'ioH, whiuli in 0-3 nor cent in (ilosod by a imrk thi'ougli
higher than tho value obtained in rawnt wliiuh tlio tubi'H A and K
direct dotorminationt). |iaNH, This vi'Hwol i mir-
Tho motlun] IH not a goixl one fordotort "- 1

ing tho latent heat of water, sinoo tlio oo''

tions depend on tlio diflovonco of tlio
volumoa of ico and water.
(3) BUNHUN'S CALoiionwmi AHD
MODIWOATIONS. (i.) Description, A
ddaal tost-tulo A is fused into a Inrgoi
bulb JJ, aa shown in J^jj/. 1. Tho bull
furnished with a glasH Htoiu 0.1), whioh tc
atesi in an iron collar 1). Tliiti stem ia
with pure boiled moroury, which oi
tho bulb to tho lovol /I Tho ronminc
tho bulb above fi is Illlctl witb
water. A calibrated narrow glasm t
hirniahod with a millimetre scale, ia
into a cork witli fine sealing-wax, and
passed through tho mercuiy in tho ooll
and made fast in the mouth of tho tub
so that it becomes filled with
mercury j

by adjusting the eork in the mouth r

tube QD tho extremity of tlio mercury c<

in tho scab tube S can bo
placed ni
'convenient point, By methods, ivhiol
described in moat text-books on pro
physios, a mantle of ice ia formed arouii
lower part ot tho tube A.

that the density of ico ia not n. constant waa duo to tho Jiiucn lower temperature at ivl*J* "^ * *
quantity. mantle was formed ill his own
nU 1 "'
Tho expori mental ovitlonoo on this point is tlieroforo made Home (leterminatiotm willi
summarised below fnwon by means of alcohol at -5 to ~ 10" aa fff i ^' 1
1 "'

briefly :

ant in tho manner devised by HiniHon. '1'hn ni <-j Ml ' J '"

Nichols reviews fliti work of previous invesli- nicnts nppemvd to indicate that the maiitl(n fa cr* 1 "'
' 5

on Iho density of ico and iksodbwi life own 11 '

by tho II.HO of alcohol at -fi" to 10 were li-u** el*' 1 '-
experiments, Ho concludes that tho dotiftity of ico than those formeil by means of <,'Q a nml iil-l***'* '
determined by weighing in petroleum, is -70 J by at leant 1 part in 1000, and further ill** 1 *' n<
innntlcs, '

O-niGlS + O-OOOOO, This result agrees with (ho mean LJl1 v

density t y i M
of the latter mantles decreased in [

valun deduced from different methods by Pliieker this amount after standing -4 honr.i in an fto- '** iim"

niul fjoinslei Ko|j]i and Bunscn (foi a Bimihu

1 1 1
variety Tlio use of and ether resulted in a vi.]-y v* l l*'''
of iuo) to four ])!nec3 of (Icnirnals. formation of ice, and tho mantle, when a eovl'nJ'* H '*' fl
His pspntimonta on tlio causes of thn variations was reached, invariably became filled with n m^l- ^ v * "
density of artificial ico wore not completed. Tlio of (inn oraoks.
method wns to freeze Iho i mantle around tho inner Vincent* later took up the mibjcot, anil ul"** ^ M '

tube of a calorimeter by pouring in a mixture of CG a i* "'

vestigftlod the coefiieiont of cubioul expatiHiou
1 1
> *"
and other. Tho unfrozen water was slmken out na He prepared tlio ieo by means of iifree/ini' nii X *-' ' *'*'

and appears to havo obtained a different cl*' llf '^'-v

for each sample prepared. Table II. fiumtHnr' h5 *" ;t '"'"
his rein]l( ;


His values for the oooOioiont of oxpansicni f (1-n-

above samples are consistent, and no uoimt'tnl i< HI
between variation in density and expansion I'JIM I" 1
traced. Vincent's mean valuo for tho dcnui L.y i"
I part in f>000 lens than tho mean of Iho ic.MiiLlri <>f
Pliloker and Goissler, IJiiiiHeii and Nichols.
Tlio experiments of Lcdno in IflQG Hiigj^"*'* t<*n> 1

caiiRO of the variations itt

density which liml IM-I-.U
observed by pro v ions workers, Ledun look *>x( r L mi'
preoantlons to got rid of alt triiaea of dinK( tl v* <!
air in (ho water used for manufacturing Lint l'n
nain pics. Ho condonscd tho Btonm from l.x>ilii>}j
wafer under oil (o obtain air-free water.
Tho results of these cxperimontaiindiealiMl Hint
tho density of ieo at wan not Icsa than O-IHV^.
nnd as greater oJTovts wore made- to remove 1-i^ifi-n
of gases tho values obtained for tho deiiHlt.y 1*1-
creased. Ho concluded that tlio density i>f MJI--
free ico at would probably bo 0-0170.
It in of interest to note that Ledno consUlaivi I lutt,
ieo made from water which haa boon merely tn tllt-il,
Jim. 2. as in tho case of tho BniiHon calorimeter, tttill i *!*!-

tains about 1 o.om; of gasper litre at

as possible, and tho adhering water
oomplotoly pressure.
frozen, the remaining space- boing then completely Anotlior possiblo oaiise of tho variations in
filled with
moKiury, Tho weight of tho moi'Oiiry,
istho strains eofc up in Iho ico block on form IL -than
together with (lint of llio ieo, gave the data for tho and which disappear in the course of time.
compulation of the density of (ho ico mantle.
Although llio results were consistent among them- (4) JOLY'S STEAM CALORIMETER, (i. ) f/V/f-
KtAvca tho iilwolnlo value was
siiliHcqnontly found to
Method. In ths atoam calorimotoi doviuotl |

wins on account of thodeformalionof
tlioglass Jo!y in 1880 tlio lieat necessary to VOJHO
"10 weight of tho contained
mercury. tomporatura of a body from tho air tomportvtii r^
""nifjlit tho (liHoro)ianey amount- to 100 is measured liy
dotGi-minlng tho
"wcmi IIIH vnliio and Iliinson's
s. Me,, Ifl02, xv.
Proe. Roy. Soc., 1880, xll.

of steam which must bo condensed into water becomes at once filled with saturated vapour.
at 100 to supply this heat. Condensation immediately begins on tho sub-
The instrument,
especially in ita -differential stance and the resulting water is caught in tho
form, lias been found very useful for special pan, weights being ridded to tho other pan of Hie
purposes. But tho experience of moat users balance so us to restore equilibrium. During
indicates that tho condensation method is tho process of weighing, the utenni is
more troublesome passed
than tho method
to use through very slowly (by opening mi escape
of mixtures for tho
determination of tho tube loading from tho boiler) into the ciilori-
specific heats of and liquids. In the
solids motor, so as to avoid disturbance- of tho pan.
hands of Joly, however, tho stoiun calorimeter After four or five minutes tlio substance
has produced data of fundamental has generally attained tho temperature of tho
concerning tho specific heat of gasoa at constant steam, and tho condensation is completed.
volume. The pan then ceases to ino'renso sensibly in
One disadvantage of tho condensation weight, and tho equilibrium of tho bahnoo
method is tho fact that less than 2 is maintained A
milligrams permanently. very slow

mo. s,

of water is deposited per calorie, and conse- increase of three or lour milligmmmcH pin-
quently it necessitates accurate weighing. hour (duo to radiation.) in, hwvovor, notluwl.
(ii.) The A2)paratua, Tho simplest form of Lot 0j bo tho temperature of tho Hlcant and
Joly's apparatus consists essentially o a Ij its latent heat. IE JH tho imiraiHO of
steam chamber of thin metal in which is weight tho quantity of heat given out by tlio
suspended from tho arm of a balance a small condensation is wL, and thin is expended in
platinum pan (Pig. 3), carrying tho substance raising tlio Biibstmico and tho pan from <\ to
under teat. ftj. It Wbe tho weight of Iho nn!>ntnnao, and
Steam can bo turned on to this chamber, s its specific heat, the boat acquired by tlio
as indicated, and escapes through a pipe at tho subatanco will bo Wa(fl, - (3,), and that
base. acquired by tho mtppoHing pan will bo
Itessential to arrange tho inlet valve so
k(0 a
~ OJ, where k is tho thermal
capacity of tlio
that tho stoam can bo admitted rapidly for pan, that Is, tho quantity of heat noccasary
reasons which are explained later. to raise its tempera-two 1" 0. Honoe wo havo
(iii.) Mdlioil of Experiment, The substance
is weighed with air in tho chamber and tho
temperature- carefully noted. Steam in tho The quantity determined by
fy is
meantime is got up in tho boiler, and is observation, and tho tomnoralm-o in found

suddenly admitted, HO that tho whole chamber either directly, by a, thormomulor inaorlod in
the initial stages of tho
the steam chamber, or by moans of liegnault's glolmles on the pan during
is somewhat counter balanced by radia-
tables and a reading of the barometer. experiment
tion from tho steam to the substance.
For extreme accuracy a small correction is still
necessary. Tho weight W
of the substance is found (5)
in nir at and the weight to is found when the
substance and pan are in steam at a The weight .
of strain per cubic centimcit.ro at 100 is littlo more the differential
form of the steam calori-
than half that of air at urdiimry temperatures; for meter the correction for the weight of stcnm
IH eliminated. In this
this reason tho weight to ia greater than tlio weight displaced by the pan
of a volume form (Fit}. '<!) two similar pans hung in tlio
of vapour condensed by excess in weight
t> of air at 0, over tho sumo volume of steam at 3 ',

steam chamber, one suspended from ouch

where u is tho volume of unbalance and pun together. arm of tlio balance HO as to counterpoise each
Tho difference of weight of n, cnbio centimetre of nir other. The thermal capacity of Iho puns
at C. and a cubic e<mlimoti-o of steam at 100 is
can bo made equal, so that the term with k
00011116gram, according to Kcgnaulfcj hence tho
as a coefficient docs not appear in tho equation,
correction to ho applied to w in OOQIiJiOv.
and the radiation error will alao disappear,
This correction being applied, tho weight of aa i.t will cause equal condensation on the
water condensed is determined, but it must be two pans.
remembered that tho weighing is made in steam; Tlio chief use of the differential form is,

and, if extreme accuracy bo desired, it is still

however, application to tho ealorimetry
its of
necessary to multiply by tho factor 1-000580, gases. -Fur this purpose tho pans are roplaecd
in order to reduce tho weighing to vacuum. of copper, ono contain-
by two spherical shells
The actual weight in a vacuum of tho water and tho
ing tho gas at a known temperature
condensed will therefore bo other empty. Tho spheres are furnished with
small pans, or eatoh-watoru," to collcet the
water resulting from condensation. Greater
so that s is determined from the equation condensation occurs on the sphere which
contains tho gas, and tho excess gives tho
quantity of heat required to heat tho
Reforonco should bo made to (0) of the
mass of gas from 0, to E This determines

article on "Latent Heat" for values of L. tho specific heat of tho gns at constant volume.
In order to avoid the condensation of afeain on the Tho great advantage of tho differential calori-
suspending wire, where it leaves
the steam chamber, meter, is that any source of error common
it passes, not through a small holo in the motal, but to tho two spheres is eliminated, and the
of planter of
through a small hole pierced in a plug gas or other unbalance enclosed in ono
1'arin. Without tlio plaster the steam condenses on of them merely bears ita own sham of orror
Hie metal and forma a drop nt tlio aperture through
and not that also of tho containing sphere.
which tlio suspending wire passes, and destroys the
Thus tho effect is practically the samo as if
freedom of motion of the wire and pro vents accurate
tho gas woro contained in a vessel of zero
weighing, With tlio plaster of Paris ping no snob
collects, and tlio weighing can bo performed
thermal capacity in the single steam calori-
with accuracy. In liifl later experiment*, Professor meter form.
wire around Tlio spheres employed by Professor July
July placed a small spiral of platinum
the suspending wire just nntsido the aperture, and woro of copper and about (H em. in diameter,
by passing an electric current through
Iho spiral, tho ono containing tho gas being mtido to
mifiioicnt heat is produced to prevent condensation stand ft safe working pressure of about !lf> or
on the suspending wire in tlio neighbourhood of the 40 atmospheres. If at the beginning of the
Hcsidos accuracy in weighing, a point
experiment this space is filled with air at about
of prime importance is the rapid introduction of tlio
22 atmospheres at 0, tho pressure will rise
steam at the beginning of tlio experiment. When
to about 30 atmospheres at & In ono ex-

tlio steam first enters the calorimeter, partial con-

tho weight of air contained wan
densation occurs by radiation to the cold air and Iho
walls of tlio chamber. So mo of the condensed 'i'28f)4- grams. The condensation observed as
globules may fall upon
the substance- and lead to on duo to tho air was 0'1(5217 grains. This
error in the value of s. If the steam enters slowly required a correction to compensate for the
this error may be large, and it is therefore important difference in weight of the spheres. Tho
to fill the chamber at onco with steam. This corrected value was 0-11029, tho range f

necessitates a good supply of a team and a largo

S -0 L ,
temperature, being 8<l.'52 0. In a
delivery tube, but when tlio chamber is well filled series of six experiments tho moan precipita-
with steam a very gentle after/low suffices. If tlio
tion per degree centigrade- was 0-018004.
mijiply lio nut off, the weight of condensed

diminished. This arisen from the

Tho following corrections are also necessary :

"T to tlio colder walls of the chamber, (a) Correction for tho thermal expansion
>m he again turned on of tho vessel, and tho consequent work done
(lie weight
J. .Toly, Proe. lion. Soe., 1880, xlvll, 21S.
'-' of condensed !
Phil. Trans. A, 1801, elxxxli. OS.

by tlio gas in expanding to this inercnsud 0-1721. .L''or carbon dioxide, tho tsluwgo with
volumo, pro.smti'o is shown by the following table :

Correction for tho dilatation of tlio

Bphoro under tho increased prossiiro of the
gas us tho toiiipomtiiro rises.
(c) Correction for tho tlionnnl effect of
stretching of tho material of tho Bphoro.
(Wires arc generally ooolcd by midden exten-
sion, but tho cooling of tho copper in thin case
is too small to merit
(d) Correction for displacement or buoyancy
arising from the increased volume- of the

sphere, both in tlio air at 1

and in the steam Tho moim result of tlio on
at fl a ,
hydrogen gives a HpoeHlti hunt JM02.
(c) Correction for unequal thermal capacities ((1) ].)K\VAK'S Liyirnt Ant AND HYJUHHIUN
of tlio spheres, GAr,oiUMKTKji. Dowur l
IIHH d^viuod a (idling-
(/) Reduction f tho weight of tlio precipita- motor based on an aiuilogmia jiL-iiuiijilo to tho
tion to vacuum. steam calori motor in whioli ho umploys 0110 nf
Professor July's cxporimonta uhow that in tho liquefied gascd ns caloiiinotric Biibatdiico.
tho case of nir and carbonic acid tho sp-ooiflo Whilst Joly'a calorituotor doptinda upon con-
heat increases with tho density, but with densation on a eiild object, DOWIU-'B oalorl-
hydrogen tho opposite seems to bo the onse, motor dononcla nn the evaporation ns a immns
.For air tho specific heat at constant volumo of absorbing lioat from tlio hot
at amoan pressure of 10-01 atmospheres, and
a mean density of 0'020. i, was found to bo
'' JW ' it(K ' A> Ixxvl ' 3 " fi ' K v '

of of gs ovnpo rated, ho fi() u.o. capuwty,

weighing tluuinnntity a long narrow tulni U, projecting aliuv*'
determines llio volumo of gas given off H1
tho liquid whiuh, of cmirao, is at its hoiling-
month of A, mid hold in its plftu l>y
loosely packed cotton-wool.
From tho
point. Now tho choice of Ikiuofied gas to IJB
of thin narrow tube, either beforo

substance is ilotsr-
employed as calorimeter a hrani'.li tuho M Js (.ulti'ii
mined nminly by two considerations : passing out of A,
ti cimblo tho volatilised KIIH fnm)
llm '"!
(a) Tho qni5iitif,y
<>* & f?
[v()T1 by
mtnivur 1

of one imlono motor to be collated in tlio I '.

evaporation on tho absorption or other Hiiitablo liquid. '1'

\viitar, oil,
of heat, and M1
of tho projecting tulm (!
range of tempo mtu
J'ho re availul>lo extremity
tost- tube 0, to contuin tlio of
is cooled.
throiigli whioli substance attached by a
is (tin*

Tho table below summarises tlio data lor oxporimenled on,

flexible rubber- tubing J), tlum Eo
so mo of tlio possible gnses

movablo joint, whicli

TA.BLB IV bent HO tilt n '" an In
tho small phiOOH of HiibHl,f
uontainod in (J into ilio 'i.
motor, ami \vliioh iifl

assumes n juiititiuti
Homowliat lilto that in
With cavii it in ]>Hnllil
iv mn^lo pii'iHi
at a linm f
iiitit II, lint nil ini]>rnvi'it
of thin I'l'doidiwli) in olic
In it, I in a wir
tyj,, it., I

tho eorlt (J, HtUnl into

(ho jiioiilli
It will bo observed tliat oxygen givca off through
Icst'tubo ullnohed by u lii'iiuoh thi'oiiuli thn
13-2 o.c. poi- caloric wliilst otliyleno gives l.'[,

riibbtir luljii to llm und of <!, na lu-fon-.

Hydrogen gives off 88>0 o-o. tmd is
only 7. Uii! oud of tlio vim ]' in it lumlt, by whliili un |

but for the (not

particularly advantageous of tlio liiibHlamio. fit a tiinii Dim In; jiulli'd nf
that tho manipulation ia diffloult. Although Whim no other
dropjied info Jl.

nitrogen is ft littlo bettor than oxygon,

it is
made, tlio pi>rlioiin of iniillor

preferable* to use tho luttor for tho following at llm lcm|iorftliiro of (ho room ;
but ivlu-u I

reason. The boiling-point of air is liolow tliat tempera In ren ftro reiniii'iid inilinlly, n viimiiiin. \
o oxygen. Even if thoi'c is no tayor of cold U, nontniniiiK oithor nolitl earlwiniti ucM. l.i

otliylouo, air, or otluir !>, oan bit pinned m> n

envelop tho tcrtt-liibo <! or (!, or if hiHb<T l<'ii| j

tnrc.'i nr-o tlm ftiirrouinliriK vcNsii'l niu

filled wllli tho vapimr of wtildi or nlhri' lEijiiUi". 1

Now, when a (jiiimtity of liquid iair linn lici'ii in

(oinf! volalilinivtii)ii for it (tine, an llio nilvngcH nvi
ntim movo ((iiintdy than tint osygon lint liuilinjj;-

MUDS Hlightly. Two jiointH rciniirn

uf lliin ; lii'rtt, tho
of (]w liijidd air
ono JHtrlrs of oxpm'imfiiilft j no.xt, tlio ]-i-vi'iil U
n. teiidonoy for tho oaloi'iinotw H to "mink In
Bomo of tho already voltitilimid KIIH.
Jli i nrO
oxterior vessol A nlioiild bo filh-d witli n,
(pmntity--Hi)mo two Hires of llijiild itii-, W
tainiiif; n higli pnrociitn}(o of oxygen, mid lliiv
oxygon or gaa on tlio aiirfnco of the liquid
motor ilflclf diioiiUl l)o filled with ttomo o[ llm
oxygon, tho air coining in contact with it fluid. Thin will iniiinttiin vory nlosisly llm inni
through tho ncolc of tho calorimeter would tomporatuvo rcquirnil. When any "imnldiiK In
still remain gaseous; but if liquid nitrogen. Hoonia to bo tlm cnlnrlmnlcp nlnm
laltiii(( pliioo,
is need tia (lalorimotrio substance, air, being
omplicd and fillnil unvw from tho lururr iln*
heavier than nitrogen but having a higher Tho Uibo botweon tho oalorimotor and tlio n'i r*-i
boiling-point, would, in falling down tho nook fihould bo of llio mo of wido quill In bin jr. nil
of tho calorimeter, come in contact with tho lower end uliovild bo BO lUTAiiged blitw llm m
cold gaseous nitrogen ftnd bo condensed, of tlm liquid in llio collecting \mw\ nn It) |^i<
roHHlUnt prcSHui'o. With miuli preoauHoim,
It consists essentially of a htrgo vacuum limy easily bo obtained correct In within 3 |ii*r.
vessel A (Ftg. C) capable of holding two or Tho instrument having boon Hot up imc.l
throo litres, into which is inserted the culorl- with liquid air, an oxportmoiib is cmiuli
motoi , a smaller vacuum vcsaol
U of 2fi to by tilting up tho littlo toiit-tiibi),

cooled or heated, thereby dropping into the stopcock leading tho evaporating hydrogen
calorimeter a portion of nny substance previ- through tho upper part of tho apparatus.
ously weighed. Tho substanco in this way This arrangement Lima charged only needs a little
falls from tlio temperature of the room to
liquidnil' euoliod in
every ono inul a Imlf hour. Tlio
that of liquid air. The heat given up by it
liquid hydrogen vessel will not need ronUmi wiling foriiL
volatilises somo of the liquid, which is curried least four hours. Tlio level of (lie liquid hydrogen in
off by tho brunoli tubo and measured in tho tlui calorimeter (loos nut full 1 om, in MX houivt willi

graduated receiver ]?.

Immediately preceding constant uso. The bulk of materials added roughly

or following this observation, a similar compensates for tho volume of

experiment is made with a small portion of tlio- liquid
hydrogen evaporated,
a selected standard substance, usually lead. His important (hat thin l*vol
should not materially dumgo,
The quantity of lead in so chosen us to produce
since, of tor striking I. hi; shoulder,
about tho same volume (if gna in tho receiver
bodies move more slowly, ara
HS tlmt supplied by tho portion of substance
deflected on to the cold wall,
experimented on, ]}y this means tho circum- and low results
stances of tho two observations are made aa lire obtained duo
similar aH possible, and thereby many sources to tlio longer
of error arc eliminated. cooling of the
In 1913 Do war * further developed the vapour lii'toro

method aa to adapt it to tho range of

BO being inimenml
between tho boiling-points of in ,Lho liquid
liquid nitrogen and hydrogen: from -190 to liyth'ogon.
- 2513 C., a ran
go of only 57.
Tho liquid hydrogen calorimeter ia a glass
cylindrical bulb vacuum vessel (Fig. (t) of A
fiO o.o. capacity, silvered, with cm, slit. On -J-

tho nook ia sonlod a glass tubo B. This

projects through tho brass coned fitting cap
If of ivn ordinary slit silvered vaeumn vetisol
in which it is
supported, side delivery A
tube, provided with stopcock I), is scaled
nofir tho top of B. A short length of rubber
tubing on the neck of Iff makes a gas-tight
joint with B. To minimiso splashing, and to
reduce tlio impact of tho falling pieces, a thin
strip of German silver or lead K, bent out
neat the top into a shoulder about 1 cm.
square, stands centrally in the calorimeter A.
Tho strip is out from a thin tubo of about tho
same diameter as tho calorimeter neck. A
short length of tho tubo is loft above tho
shoulder, and supports tho strip by fitting
loosely into tho nook of A. Somo such
dovico ia essential in tho uso of this form of
tho liquid hydrogen calorimeter.
Tho calorimeter in its turn is immersed in
liquid hydrogon in tho supporting vaomim
vessel C, tho nook of tho calorimeter being
8 to 10 cm. below the liquid hydrogen surface,
Tho vacuum vessel is only slightly wider Via. 0,
than tho lower parb of A, and is provided
with a eoncd cap V, whereby it is also Tho isolation of tho calimmelcr was nneli (ho.t 1ms
than 10 o.o. of hydrogen gnu WUB ovtiporalcil from it
supported and completely immersed in n
wider vacuum vessel Gf containing liquid air. por miimto, The wliolo n])|iaraliiH. w Bujipotlod 1)fi-
twcon tho cork-lined spring jnwB monntwl on a heavy
G is also with a brass coned cover,
metal bnsc on wliioh the milor vnoiiuin vessel rest*.
fitting vacuum-tight on to tho cap F on 0,
Both caps are pierced by two thin tubes, one Tho cooling vessel H
is eonnootocl by an

for fitting on to the filling syphons, the other, india-rubber tube to tho top of tho calorimeter,
bent at right angles, serving for connecting to It consists of nn ordinary oylindriaal silvered
tho exhaust in tlio case- of tho liquid air vessel, vacuum vessel, 20 om. long find 7 oin. wide,
n.nd in tho case of tho liquid hydrogen to tho with a central axial opon tubo 1C mmliwl in
Prac. Eon. Soo. t 11)13, Ixxxix. 158. below. This tube passes throngh the liquid
in tho viuiuiiin VOHHO!. It has bho same cnipm'Jiry bnok pressure being fiilnl to ncoi'<l^ 1 "

diainott)]' bolow us tho iik tube of tho cidori- L'fiiilts. At leant 16 seoouds arc allowed tor noil*"1
; ' '
on **'

tho gas given and even longer, in s.omo

'Ni'iu* tin! tup i>f the central tube a
muter. rig

Hi i In tubu J, of about Iho name diameter and ivith budly conducting bodies,
woro t"" %H

Hoimi II (MIL lung, servos fur tho introduction far us iKiNsible tho materials imod
of tho woij.<hod piuct'H of material, which arc n this forms of aplieros ahnut 8 mm. diuim^*' 1 '

the irmuM ^V
nil iiotiltuJ proviuiinly to Iho lorn \i am tu ro of Intlio cnso of liquid bodies,

liijuid ail', and tlum fall on to

a thin metal 'irst coolod by liquid air. liYcriiionll
(tine I* Jilting tnbo K, where they
loowily tlio wore fiw.uii into acilid cylinders in thin
roiimin ahimt Ifi minutes. J, in mipportcd by

and pieces cut off after removing t-' 1
HV** 1 "*'

lining hitij^d lo two thin nbonilo rods, L and flaas mould, Tho mctallio vnatorialH
iM, llxod tn a briiHH Fitting comoiitert on to u Homo cases fused into buttons of donvon J* 1 ' '
* ' lt '

Urn tup. Tho r<id is not fixed direotly to

J'j rt'oight in an oxhauated qnurtai tnlio, '

Mm diHii IniL Li i iv motal ring It. From tlio ead, howovor, of whih many piccow \v
riiiH -it two thin vortical stool wires are was cut from rod, and smbHcquoK
isimiiouUsd freely In two 4111 Iho isircuni" Hqucc'/.od in a small sjihorical
f tlm pun hwlow. Tina rod and tho Volatild binliea wore weighed at ib l

ritif* t;iin bn given a vortical motion tomrjoraturo on a lif'lit Gorman mlvw

by n oninh N in th fop Jilting, thereby Hitpportiid by a thin platinum
wire HU >u ti t* * ]

tlio and i'<ih*aiiij fchn picuo of 'ruin tlio balance pan about 2 <im. nbi)V*i t.
tipping ]iiin
luatiiriril, w-luiih Uicn fullH frcoly down into ,ovol of liquid air contained in a wido (ti' *J
thiHiiiliiriniHlc*!', Ahi^li viusmimJs niainlnincd vaoiium vosHol. Koine materials would ii

a (H'oHn-tutM) S, opL'tiiuf; Id tlio anmiinv malto (lohoront hullots or cant stickn, anil tli"'
wore filled into very thin-walled uylim 1 1

** '

ii|iu<Hs llllod wiili tiliai'oniil.

tlio 0011-
motal oapHiileH.
Ad lliu linnjii'inliiro
of linillti^ nitmycn,
vcittrnl tnlio of siieh
In order to obtain oonsistont n.iHiiH l-
vnHlliHi inivronlM hi U)i! iv vcsnol,
wlii-n tuniiKMjlcd Ui Hit' Kiiltinnidlci

lnvlinv, Imvo no noeesHary to employ oxaotly the HUIIIO ;pi

.Hiivloim ufl'tnt oil tin* toinpnmtiiro in this tubo nt n ooduro in teat, but with this uppuiNtt-
ri'iimumbla Uiiilanuu from tho liottinn, pravidcil Uio Ilownr waH al)lo to obtain results whioli rn r
m-tilnil luhd ho n<iL wiiii- With a InrRor piillcni vrasol 1
. varied among themselves by nioni tlum --
tlu'widUntf tliin uciil nil till 10 WUH im)it>nncil t2-2om., 'A per cent,
tn Hi fivni IHJW Uin iliifcrt'iico won unilur i) aL (ho level Tho tlntu thus obtained for tho IIKNUI HiH'^
of Itm pini. 'J'Jicao tcinperiiUirt'tt wcro mcnmircd by beats of fi3 elements at about fiO^nlm. t

n niimll livllnm llii'riiuinitiii'v, consiHUng of a d-o.o. aninmarisecl in Table V.

Inilli Ui wliiish wiitt WNibsil a Hiuall moron ry iniuunnclcr

of Iliui i-ii|)illary lulling.

'''<>'" ""> Htjiiid in Iho
hydi'iiK''" cviUMn-'fttiiiK
viiiiuiiin vcHsiil (*,in wbioli llioi'iibminelttr i iinmorflcd,

IM ctniiloyi'il In UifrintiTvol cif ubHprvn lions to rnnin-

Inin n ntiu(is|i!L(-ni llu'ongli
tlio ncok of this
cnl^riniDti-r innl tlm (MHini-dti'd iiifiimirinK lubeH
IHnlt of Holid nil- in llii- iNilnnmnlcir nwilt iw HIUH olivi
ulril, A nini|itis iirnin^oiiu'iit
<>! it I
lir(!i!-wny uoolinjj

of a jduHM tnljc 8 oin.

in iliiunotnr and 'If) ons. long, open at the
bnLlimi and pmviduil with wiflo T-pieoo at it

th(! Lop. Tho lubo in iinnnTHod to tho nnok

In wiitm- in a glum oylindnr, and in omuiler-
by n, wmjrlil im<l curd running over a
/ abnvci. 11. in thiM-ohy readily
I dm'intf tlio limu RUN in being wapnratei
fniiii Dui (Milnriniulpr j Ibis ensures that IK
haul; nriwHiirn to iuxidncHsi'1.
Ono arm of the

in ciiwii mid oonncctlfl to the ntiioiiuk

.D'on Urn cmliirimcitor nooli i (bo nllier in

provided with anotlmr aniftll stopciock mid

(')innwlH a SIHt-n.n. gafi bui'otio \V similnrly

iimiwiwd in water, Thih latter sloneook i

ohiHi'd while Ihn M IIH ovii[)i>rated during
IH titilk'filod,
TboHO tirrBiiBommifi nre noccaflnry to HCOIIVO tbo
Niinininm Impediment Ui the ovniioniting liyilrogon
whioh \* iiBiinlly ovotvcd in ICHB than 10 Ht-condB, anj
TAHLH Vcontinticd tnkos pbuic mjridly, it unnnot bo cliini-inttcd or
even ftooiii'dtnly ullowad for. Soino lieat is
loat when tho onlonmolor is raised tihovo tho

tomporatiiro of ita oiioloauro niul bofore tho

final tonipwaturo is rcuched. This can bo
roughly estimated by olmorving the raio of
change of toiiiporatiivo niui aKBiuniiifi that tlto
boat loss directly proportional
i to tho
duration of experiment and to the avcmgo
excess of tcmporaturo. Tho accurate deter-
mination of this correction in of fundamental
importance- in this method and n del ailed
dinmiKsio'n of it will ho given. It in always
desirable to diminish the lona of lumt as mneh
as possible by poliHbing tho oxtorior ol thn
tialontnotor to diminiah radiation, nnd by SIIB-
poiulinj.; it by non -conducting HiipporLa insido
a polished case to jiroteet it from draughts.
It is also very important to keep tho surround"
ing oonditiuns as constant ns possible through-
out tlio experiment, Thin may bo scunrcil by
lining a largo \vatci]'-bii.tli around the apiifiratus,
but in isxpcrimoiilN of long duration it iw
adviaablo l.o utiis an ninsuralo toinporaturc
i-egithitor. The melliod of lagfuing tho (uilori-
motor with co|,toii-\vonl, wlmiii in often re-
commended, (limhiiHlu'M tho heat IOHH 1^011-

Biderahly but rendcm it vei-y uimc-i'tohi, and

should never Im lined hi work of precision, si mso
tho had condnetoi's tuko HO long (n u-aeli n-
stemly Blate that tlio rulo of loss tl(ij>oiidn on
tho punt hintoi'y more limn on tho (< v iii]ioralurw
of tho calorimeter at tlui monicnU A moro
Tlui in toi'tiNting faot d hoovered in tins
Horion.H oljjoetion to tho UHO of la^gin^ of tlii
course of this investigation was tliat Uio utomio
kind i thu dangur of its absorbing inoiHture.
lionts wore periodic funti Lions of th atomic Tlio loant trueo of inoisturo in tlio lagging may
woight, ami tlio curve roaombled, gonorally,
the well-known Mayor atoiniu volume for the
produeo serious loss t>f Imiit by ovajiorjiiion.
about IH'10 nmdo a carofnl study
solid state. 10. o,
o tho Jlethod of Mixtures, and Iiy wldll dint
attention to detail he obtained by nmnm if
OALOUIMETBY, METHOD OF MIXTURES it a valuable noiion of thermal data, It in
(1) INTRODUCTION. Galorimotrio appanvtun only within comparatively recumt yearn that
assumes the most diverse form, oaoh typo char- any material improvements on Bogmiult'H
acterised by certain features, which adapts it apparatus liavo been offootcd.
especially fur a particular el aw of iiioimui'omtinlj, (')) TIIEOIIY or METHOD or MIXTVIIEH,,
J.i"oi' tlio ilotormiiuition (if tlio moan Ftpodfio It will lie assumod Unit the wises uniEor eoii-
lioat ot a HiiUatuneo ovor a mii^o of loiiipcmtnro mderation urn Mi OHO of Holids and H<jmdn,
or for lh (kitoniiimilion nf thn himt of oom- Tlio deturmination of lluv ^[)oui(i<! IKMI(H nf gimon
biiflUon of 11 funl tlio Motliod of Mixtures rmiuires enpeoial eoHHidtiralion and IH dealt
ifl a ijonvdiiiont uno to employ and SH
probably with in a Htiparata seetion.
tho bout Itntnvn of all oaloninotrio mothodH. Lot Mmnas
of honied body,
(2) THI-J Mi'M'inn) en-' MIXTWIIKS. 'J'ha jirin- tf:=tom})oraturo of heated bfidy at tlio
ciplo of tliin niothod En to impart bhti quantil;y moment of its immersion in tho
of hont to lio inoiwiirod to a known mnss tit of tl> (jalorimoter,
water oontaincd in a voHHtil ot known tlionnal of water omployod,
oapnoity, and to obmirvo Llio rm\ of tompm'a- 4:-= temperature of water when tliebudy
tui'o produced from wlimli djitu, an oxplained
; mimnuMwd in it,
in detail farther on, tho (piaiitUy of Imnt uu I
temperature when tlio thornwl eipil-
bo oalmilalod, 1'hto molbod in tlio tuniploHt nf Ubi'iiun IK t'ntablmliod between tlio
oalorimotrio mothoilH, but in not Urn inottt hitdy and the water, i,c,. wluni
aoonrato. ]foat in loot in truiiufcrnng dho liot toinjioriiLuro o[ water iii-uwm to
object to thp caloriniotor, and although id cull rise {or sink if the boily
bo mini mined by arranging Uial tlui transfer than the wnter).
fifi METHOD 0.1'
Now if H-..-:mtfitn upoeilii; bout of wider lw- of I'lmdiiimr (inn ID its britllonosa, tliw lujy (

envu 'I
and ami M boat of Iho body ]ioi'|.i(i!in
of tlio glafs wnlls abovo tile Hiirfaoo f
(, .~n|iomlic '*
i t) mill T, wo got
water, Uiodilliuiillyof aHecrtflmingits ln;ii.l.
inwl of kcopinjj it cotmtimt.

Under (iortain oircumstiimses tlio nsii

walot' or othci caloriinotric fhiid bncomuH


]>nu!tioiiblo. lU'-iKH) sonio invMtigat

Thin tfiveH tlm
for Iho
(raloriinoU'io equation eniploytid as caloriinotcr t-hiok-wallod iiH'liv""''
very simplest form nnd truatcd to Iho high cocflioi^nt of !I*' L|

iimtlioil <if luixliiL'os ilk itH ;

mivonil ooiTontions aro conductivity oE tho metal to cqimliso
for pmcilitsul *
working '*.**
** '"
n in T.H miry. lo])i|ioraLure. Tluwi siicc'ialiscd forniw of onl<
(i.) Fur W wo numb milwlitulo another nmlora aro dose ri UK! Inter. I

W, midli that W|"W-l-imm oE (ii.) The, Slirrer, iStirrot'H vary eoimidonv' ''.

, Imivmg llio muno heat uajmoity ns eon- in diMgn according tu tlio flpcciflo purp*
Htirror, mid nil purls \vhono for whinh they aro rcquirod. A typi*' Jl ^

in * ""

ti)iM|>m'aluro IK mirimmly ud'eolud : fiu'in IH illustrated .I'ii;. 1 ; hoi'ts

I.e. if /i
-: miisa of tnohi! roaorvoir, Hlirror, oto., K(irt\v ia employed for utiiTing Uin oimto***
mol.nl TIlO VCHHol in (lOllHtl'lHllCl! Wltll llll CO -
mid fl[
-.111(11111 Hpeeifii! heat (if this
T mid t,
(H'oHa-sdolimi, Iho Hlirrur
buing (lontaiucd
tiihii which in connected
\v-w -i-/;, with tlio main tube at tho
top mid bottom. Jly this
jind in thin <WHO (iiwiion hocomod
arrangomont it if) iuwmblo to
miaiiro a steady circulation
Hr= 1 ' ' '
'^ ami
jyiV-T) tliroiigh tho caloiimcter,
it IB tho
iidviHiiblc to direct
HniiHitimoH tho subHtunco has tu bo supported
strawn oa nmootlily us ]ioa-
in a ixinoptiusle, and If wo put ~; water-
siblo by suitably curved
tlio limits of tho oxyoriimmt, wo get
In tho design of fuioli

.. M flalorimolor partioulav at-

' ' '


"At(-T) tontion must ho givon to tlio

for tho csirmilH,l-i*n
WJioro tins roHiiltfi aro to lio as aoourato ns provision of wido passi^ea
a nun'o compHualwl
of tho wator and oaro tiilcon to avoid an fjii"
jiuHHllilo, oqufttion (*!)
(w jxiHHiblo dead apixces. It
might bo re-
Cuvii), In luliliUoti to Uio iniiKiiitudca already
marked mothod of stirring him lM-t*n
wu Imvis to tnko ucoount of Uio
Ihunniil onpacity of Iho tlinnnnnHjtoi'fl, olo.
(inmd to bo
tliat thifi
tlio miml rolialilo for conipiirln* m
untlis for morciiry thoi'moniotom mid in nni*'li
In iiddiUfHi to thcsso tiomMiliinift doiumdont
Hii]iorior to a plain nerow in a voaaol of liq.iii*l.
tin Uio iwituro of thu vnrioim pnrla i)f tlm
fleirrculion fur Whito adopttid a Honiowluvt mmllar
i t'*tnn
tlni]'^ in tlio (MiolhiR
H-ppuiiiliiMj l-lit*
did 11 ri motor to of oalorimoror for his oxporinionlH oil
tlui low <'f linut fi'uni l;ho il

heat of Hilioatos at high tonijicratii rt>n f
TKK M.KTUO1) Tho oliargo, contained in a platinum
orimi lilt*,
(!) MdllKUN
Al'l'AltA'l'Urt ]'(>
waft dni]i[)od (Hrcclly from tlio fiirnuoo into t 1 1* ?

ot 1

,M!XTIIIII:H.- --'i'ho (tunijilold on Kit niqnmsil

for ox])irinn'ii(H wlHi Uio inntlitHl of tnixturcH
It will bo oliBorvod from Fiy. 2 thal> l-lu>
cniihihlH of lli following nloinoula, wliioli will
IID uoiiHidovt'd individually :
covor W
in in actual contact with the \vn,tor

ami Urn dovico to onHuro tomporaturo o([iiililit'inin. Tho tin-

(i,) Tlni oulnrimotrio vossol
niilar Hpaco imdor tho covor pormits tJo \vi\t * r >

fin- mixing Uio auiitoivte.

oaltmmolor. to vary Homowhafc in mnnuiit withtmt <iv<i'r-
jtmlait oiuiloulng Uio
(li,) 'J'lici
wliiln tin*
UiotiiKHiiotoi for nioafluriiiK
Uio flowiiif* or failing to wet tho oovor,
woighli of tho oovor (70 gnu.)
proven*-** it
iDinpwniliira i-lwi of the wator.
or cooling from bring floated out of tho place wlmn II it-
(IvS)ninilltmoo for liwilinfi

lioiit (lolormhwllonB).
water IH high. An apnroximato prclhniiiti r,y
tlui dlmi'Ko (in H])i!ii5ll
of wntor is of
ftdjiistmont of tho amount
ti ivs.*- t

jj(0) (JAMH11MKTMII AMD HTIItRXII. (I.) '/'/ifi

<}itlorim<'i(!r,-'M\Q milorimolor IH usually mndo uoooBHnry, and in very easily l'Iv*ij-
df pnrn wnppor, nlolcol plnfoil ivistt jitiliBliod HO
oration through tho joint is about 7 mpr. mi
lit n.Hlinso nuliiiUtm lo a minimum,
hour, which is notlikoly to prodiico npprmilft'l,!^
orrnr. An oil seal gives very littlo troul>l(,
Tfm HO a viMnintn frtokul onloi'l motor in not
(if IIH ft
and is uaotl whoro maximum noonmoy ( rt
to reeoimiLtwk'd for oriliimry work. Undoubtedly
Jl IH ponHililo \>y inonm of it to rorluoo
Uio mngnitmlo
of Uio QooUng rule, but thin mlvmUagu i more limn '
"Somo Oalorlmotrlo Apparatus," I'fw-
llm (llBudvimLngOB ot this form Roc. 1010, xxxl, Mo. 0.
condiKsfcion, convcisUon, radintion, lltlt
Alwoi'li.ul muinLura if Ui surtiico fi hygroseopio. liy j "
Tlio tihjcct of apnlyins the cu >< *"

A 1ni'iiis!inl (!ot)]ii-i-
mirfiuii 100 mil. mi\um in nn-n, ovaiioniliDn.
Batumi CM 1 I* **"
will iiliHorli 7 ininifeminK nf water in n ing con-ctstkin in to cliininnte lliis lioat
wlnilsl 11 sia-fnco fi'om this linivl result.
nLiiUMplutrn nt ui'diimry Iwnporatuvc,
of 'tho name ni/.n and imdor tlio nut ncocssnry *'*
of [loliuhttl niokel Whilst in practice it in ' *'
HIUIIO <iinilitiotis will uot absiorli ns imioh
an 0-1 1 t
study tlicso effects separately, it tni^ht lio i
to provide 1 *
milli^nim. lli.'iini; it in nlwaj'H advirtiiblo marked that under ordinary cjcmtliiiniiH i-l*'
tlui c'liloriinulc-r with n lid, altliouKJi il will
bo found *'*
greater portion of tho
heat transfer i* cltu 1'

Unit In tlio i:iil(rimiiU>r wll'coliviily tlm addition t

convection and air conduction, tho two tojjfot-l

f 11 III compliea-tiw llio conslrnutma.

odnsLltuting jthcmt 80 per cent
of llio tnUil-
with t!in Hiinplr-Ht form of appamliiFi a
'It is advisable to roduco tho tnmn[ir
uiilndmMid imjmivwiKsnb is jmiduced by providing "*
{if lltin iniiUil havJujj ft slob to allow of its thermal conduction through tho mippnrls* "
11 liil vory
tlio calorimeter to a minimum HUKJIJ it 4'*"
of llio hot liody
lioinn nsinovcul for Uio inlrod notirm
Htitutes an iincontrollahlo Hunreo of ornn

willwiut dwliii'biiiK tlw t kormoiiHilor.


rl ~
TlIK 'J'IM'3 OAf.O-
jAlllti'iT HIIIUIOUNDINC! Consider, for exam pip, tho CHHC uf a L'iili*
motur supported within tho onoluHtirn <i*
()S) l
uiMKTJ'SH. While (luring ourwo of his 1 tlio *;

sheet of tioi'k or ruhhor. When thn csalurjnn'1

oxlondud rcHoiivt sites on tho upeeifio heats n **^ 1 1

tho MliisaUw htta dovotcd nnwsh nttontion to and enclosure aro at two steady Uimpwati^
the ilemgn of tho tho heat tmnsfor is, by tho IIUVH of i!omhust.L< *
1 >*|.
proportioniil to tho teinperrttnro
iHiloi'imutrio difiure'imci ;
p ii rat us oin- however, tbo toiiiporaturo of tho citiUirim^
is changing rapidly tho rate of tranter in
*' m*
jiloycd. 11 "
()no form i if oven a|)pi'(*xiniatcly pro]>ovtioiial to tlio i*'*
juukot for main- poraturo di^renco.
lainiiigaomiatant The following ia a diocuKHion of this mini 1

*"* 1
011- of error by Dioldnson :
aroimd Sinuo t'hc conduotivity of Hiieh miilmin-H*
tho euloriniotcr ahvayn small, compared
with that of
which ho has do- mctallie slutets in contact with thnnu
Heriberl in shown tomiioiuturos of tho siirfaticn mivy
Im lulc^-'Ji
in Fig, (i. Tlio for the pui'pnfica of this dismission, IXH njiin'ttx
i'*' !1
jaokut is ftlunvn matoly tlui namo as the measured tt!ni|)(H'iil.i
in Bootion and of the oaliH'i meter and tho junket i'OHjieioliv*'l.V.
also in top view, Tho (liHtrilintion of loinpoiuturo in Hiich u hi-y t-r
at which heat in IntwinM l " 11 ^
and tho nito
oalorimetor at any timo may ho them lioU.'i

Wtitou is held up
in tho two halves mined from the following coiiHidonvtionrt ;
oE tlio cover, and
A sheet of malerinl of thichiioHH c, lioinulol *'.V
i n t li o u i) por Hiirfin!*'.H and a w iiiitmlly at twiinnrnl-ui-n
1 *'.
jilauo a'o ,
ho nrc Oftoh Inkon UN (t for (nnvciiitniiu'. >i
l-'ui, fi.
> I*
H .!<,

ono of the mirfmu-H *\ In llM'H

tlio temperature of

muswl lo rino from C to 0' in mush iv \my lluvt

H, hllmir iiulliiy} .Kciiii
of H uro. On leaving (llstrilmlion iti Hif'
|ittMH!i)(o UiurinnuMiOtrji 0-0'(l~e'""'), the lompevalni'ti
iiKiiuiituL-; W, water luvel. , *
]mjl)0 i| ol llm following cctiiutlon :
plato i given by
wutoi tlividoa imd
1 noross llirough tlio IV" ,
ami thon votwrna throujjli i-sln
IHi'flo upliui' i)aan]ica
Itii eiroiilftLlim ia <lirootod by
tho lowor Hpiwo,
bhu parUttons P and Q, of which (i HIIIH nearly
thn whole length of tlio limk uutaido
aa slmwn by the dotted lino. Tho a Iho thcrmoinctric conduotlvity
Whom fi In
i9 nponcd by moving tho covers mntorlal. F(') ta tho lompuruluro of HIM fwH .-*-
llioir down-tuniod onda tlion movo
at tlio oiulu of tho tank,
tlio U-iiitghs loft
Tho point of interest in llttB diHiuiBi()ii J rm,l4> H"
oovors wlido upon the stout rod T. Tho puttoy nt wliioli hoat is tawinf! tho eiilorimolw ul> any 1 1 , M .,
[or tho (laloiiimitor sLii-ror, and tho "\vholo ns tills dolcrmincs tho valuo
of tho cm iw*i v U->- f r , ,

sliu-op, arc horno on one half
oovor. -IT t.t ,itt<\
jiviikut tho portion of tho surface in qm-atlcm.
1 1
Thin linlf in oliiinpcid iirnily in place during ft" o nro each mado unity, and the alwvts BSJIR-HH i* f ( .v

other fully with reaped; lo tho fi>Uovvlnn

iiuiviiifi asidti tlio diftccontliilcd a;,
olwni'vatiiHi ;

motor oponitig-
'H tlw (tnloi'i * "1
nml tli
t'i -.HI-H.IH
1t f
Combtistion, Calorlmotiy,
Oombiiatlon, Calorlmotiy,,""'!
HfclAail of .mppmliny Ilia Calorimeter.
CombiiBtlon o Ciuio HiiKar, Boiuoln
Paper Jiiir, Airfff.,
llimt transEov Ixstwocn tlio oalfmmotor Nniilithnlono," Sei.
2 HO
il-H jaokofc may take plooo hi four ways 1

wr's Series )irf Spherical Ur>ntt tt f rtt ,

hoe, tit, p. 110,
ex,,r^ion for tho
tonipomtnrc gradient at any point
ni tho nnitonal m found;
l ably that of (ho jacket, hemx> their effect is entirely
negligible,both as regards
cooling rate and Jicafc
(lii.) Heat Conduction
alony Ihe Stirrer. Tho ateel
alirrer nhuft which enters the calorimeter should ojicl
- OH nnrxla
2^ ( 1 }'" just nbovo it in n thin rubber sleovo, which BhoitM

-a fit
tightly over it and tigdtly within n
Tho Hiirfutm ,r-1 larger wtccl
i.s tlm mtrfnoo In contact picoo coupled to tho driving shaft.
will! Ilia It ig evident (hat
culomimlnr. HI) Unit since the heat
milwtitutiiig tlm vjiluo of x condiiotivity of steel is many times
mid llin
upimipi'iato mint* for and greater (linn that of the haul rubber
a, Iho uljovo sleeve, tho
oxiircuHioii R [vi!H tlio lompomluro gradient in tho temperatures of the. two metal parts will remain
iniilMiiil inoondi.it will, the
cnlorimoter, which is very nearly the annie as the temperatures of tho
pro |irtinmil to tlio factor k for this portion of the calorimeter and tho jacket The heat
' respectively.
capacity of the rubber sleeve. Homo of which flhonkl
AH nil cxftiniilo Hlimving tho effect of tliia be added to that of the
kind calorimeter, IB infiignificatit.
of liidtvihulion of nmtorinl,
miripaso that tho onion-
motor rcslH on a uliool of obonito l o m g (7) METHOD or CALCULATING THE C'OOMNO
tliiok .

and Unit] tlio ConiiECTJON. Sumford. 'RnmioiA vrtia

toinjiomtiiro in ita calorimeter (i.)
rimiH quite approximately neuording to the relation tho first to introduce n method of
(ji for the heat loss from tho His
0-0<-0-fl- )(0 l -0,). where a -(MM 1
4i(l 0,0 ( ,
and tf, ropivMoiit, respectively, tho temperature at nny
procedure wns malio a preliminary experi-to
time, tho initial, nnd Mio final
ment to ascertain
approximately what tlio
temperature. Tho
valito of n , (ho thei'inoinotrio for
rise of temperature would bo and then to
approximately 0-001 in 0,(I.H, units. Tlieso cool tho calorimeter half this number of
Hiihstitiiteil in tho above
equation show degrees below tlio temperature of tlio BiuToimd-
that after (10 Hcoondfi tlio nilo of heat IOHN iw 2-7f>
iiig atmosphere before tho next experiment,
(<inu>H il.i llnitl valiio, nfloi Ti minutes the lulu in
1-13 .for example, let
Union tho dual vnhn>, nnd only ufloi' 10 ininuto-s does
it litimo tn within
pin- omit of \ta iinul value. If I Tem|)ora.tnro of atmosphere = i,
tlm urea in coninob with mwfi it Bhuot worn a coiiHidor- Approximate inorraso= 20.
ulilo jutrt t)[ HID ivlmlo area of tho oiiltirinujUa 1

, tho Tlio oaloriinotor in cooled to

nrriir inlriwliioi'il frniu thin iniimo
would lio (t-0), mid tlio
ovitionUy heated body tlum introclucod
n very Hcrionn oai Kuuli n <!in(.ril)iition of material
tlio maximum :

IIH horn dlHiiiiHrti'd wilt alwo havn an oH'iiot on toinporaliiro will bo upiiroxinmtoly ((+ 0) f
tho liral
uf HID imcl Itnmford'H idea wwa that tho amount of
iiiijnmil.y <!iiliiriiuo(oi'.

Tlim nhuwu hcut gained ]>y the oaluiimntor

iliHisutiHinn Iliat till no n-cond noting during tlio
Hiip]iinii nliould hit m^lij(i]ily mnull, or, BJIICO tho time its temperature was below ( will
tlicnaiiinotrio cumilnolivJty
"K/o/) ( tho ab.soliilo compensate for tho amount lost by it wliilc its
ciiiuluitlivily 1C divided by the- miooiflo Imiit (o) nnd toinjjeraturo WHS above 1. This ia n-pprnxi-
dniHily (p), tint intiliirinl Hetl for Ihoin wlionlil huvo a ntatoly true, but not quite so, owing tn tlio
Hiniill duanit.v Hpnoiilo heal. A form of minpnrt
nnd fuel; that tho rate of increnno ol lorn
Hhould Ihorofom ho wii]iloyod, in which tho mnnllfvit
pom turo
dimiiiishra very rapidly no tho heated
piwullilo iniiHH of iiiHiiliiliiif; nia-lci'Jal EH iiwd, with iho
and Uio wnloi- approach thermal : equilibrium
mnalli^t niitwllito arm in contact with tho caloriinctisr.
thufl, it may huppon that tlio riao of tompora-
'J'ho imiHH of Huoli inipjHirlfi can rt-ndlly bo iimdo
tiiro from ((- 0) to f will oeour in loss tlmii
iii'HllKildd i!oni])iii'fid wifh that of Iho calorimotor.
20 seconds, wliilo tlio rise from t to
(ii.) MuMMifta ami Npuca Irtwr.an ditlorimeter ami (M- 0)
Junket. TII ivduro OITOIVI due i tho tilmvo canno will occupy over 100 scconda,
tu it luiniiiiiini IHi'Uiniiini ('i))|il(iycd tho following Arithmetical Melliml of compitl iny Ural Lu.ts.-
uf ini|i]HH'tn fin- hix lialoniuolcr.
nn'an|(t'iLii'iit 'J'lio
A more nnouriite, but not, n curly so eauy n iiK'thml
itii])|iorlliiL; \i\\w\\ (tlnvn in niiiiilior) nni cuoh mndo uj) of oomwlion in tlio following: ('mil tlio uttliirinioU-r
of a IniiHfi riino niiMon-d In tlui liiitloiii uf Hie jackol,
Bcfvcml (lc(fi'cs b;li>w tho oiinlostiiro and lh(i viTy
nnd a luiinll ivury lip a In nit li nun. in diainolor
careful rwidiugH t inturvnlH of uliont 20 BWMiiulH
niinionti'd iulo tlio cud uf thu itonn and nsating ngninst boforo nnd uflur tho intmdiiiiLinn of tlio hot body,
ttnitiH plud'H (I'lin with a holo, uno wilh IL wloL, and tlm nnd nlno nffi-r tin; mtabliHlinniiit of thoniukl
Uiinl jilniio) un Mm Iml.lorn nf tlm i'alorim(iU'r. Tho iu liiitwcm I ho hut nnd the iviiti?!-.
Ihoi'iiial (sinidtiiiLlvit.y uf tlio ivory tips IH Bmall, and
Lot 0, f>i, (la, . . , On IKI (Im tomjuiniLin'cfi o tlm
Uii'ir lolid ruiiiifi in not uvci'O- urn,, MI Unit Mioir onVot
niilori motor nt llio lifj-inning and nt tlm wicl of jn
on IliKOixiHiiK 1'itlo inliiii mnull to biixi^iiilioant, 'J'ho
(irUicl of, nny, '20 iimmitH bufnro tho iutto-
ImiHti iidiii'.'t, they huvo n tiotiNidomhlo nmm,
ilncitinii of tho luiL.boiIy.
Iiavn 11 Ki'i'til. eiiinpiiiTil with llio
liciit (iiintlui'llvlty I'D
T.ol /, /j,
a /,,,
. . . ln IKI lh l(i]iijicnUureH for it
aiiioiint; of heat v/Mf.h tlioy cnn vfoiivo liy radintion,
]ii'rhuli) of I'fjiml (hiration after Iho IiitirHliictJoii n[
iiuiivouliim. olo, (uliuiit (I'CHIOI cahirlo
pi<r Hijiiaro Iho hot Iwidy ii|i
tit tho rAttihliflliincnt of (humml
liontiniolro mjuniid JUT ili^iro
|)i-t li'iiipi'i'atiiro illlli-r.
ci[nili!irinni liohi'iwii Dm hot body mid lliw ivntm 1

mice), Miat tlio! t'

lompMi'iituto in nl, nil limcit iiu-Hniir.
Lot T, 'I
, . , bn tho tomiioriitiircn fm- r
,, 'J'y 'I',,

vi(iiln jiftrr Iho rjilii.blinhmont of thcriiuvl
Viiliin fuinul liy i
liy lUt-Klnsnii fot his r

lurliui'li'r. ni : l, t nnd l' ui'o

virtually tho mitw. 'I'lio
IM. vil. iiro ( iihonlil nut 1m Uihi>n fmi tlio
of tfoo thermometer, bub BlioiiUl lie calculated us to ita. maxiimnii vnhio, the hd't '

accounted for if tho temperature ' H

whioli tlie hody would huvo
had buoii no loss. Tho