IAWA LIST

OF MJCROSCOPIC FEATURES FOR

HARDWOOD IDENTIFICATION

byan

IAWA Comnilttee

4 .,

E. A. Wheeler, P. Baas & P. E. Gasson (editors)

k

IAWA LIST OF MICROSCOPIC FEATURES

FOR HARDWOOD IDENTIFICATION
with an Appendix on non-anatomical information

IAWA Conirnittee

Veronica Angyalossy Alfonso - São Paulo, Brazil

Pieter Baas Leiden, The Netherlands
Sherwin Carlquist - Clareniont, California, USA
Joao Peres Chirnelo São Paulo, Brazil
Vera T. Rauber Coradin - Brasilia, Brazil
Pierre Déiienrte - Nogent-sur-Marne. France
Peter E. Gasson -. Kew, UK
Diciger Grosser— Miinchen, FRG
Jugo lhe - I-lighett, Victoria, Australia
Keiko Kuroda Kyoto, Japan
Regis B. Mifler— Madison, Wisconsin, USA
Ken Ogata - Tsukuba, Japan
Hans Georg Richter - Hamhurg, FRG
Ben L H. ter Welle - Utrecht, The Netherlands
Elisabeth A. Wheelcr Raleigh, North Carolina. USA

edited by

E. A. Wheeler, P. Baas and P.E. Gasson

© 1989. IAWA Builetin n.s. 10 (3): 219-332

Puhljshed for the International Association ofWood Anatornisis at ihe
Rijksherbariam, Leiclen, The Nerherlands
 221

PREFACE

This list of rnicroscopic features for hardwood identification is lhe successorto lhe Standard
List of Characters Suitable For Computerized Hardwood Jdentiflcation' published in 1981
(IAWA Bulietin n.s. 2: 99-145) with an explanation of lhe coding procedure by R.B. Milier.
Thc 1981 publication greatly stimulated international exchange of information and experience
on characters suitable for hardwood identiflcation, and inspired considerable debate on lhe most
desirable coding procedures and ideruiflcation programs. Therefore, aL Lhe IAWA mceting
during lhe XIV inLernational Botanic.aI Congress in Berlin, July 1987, it was decidcd to revise
the 1981 standard list. Because oU Lhe continuing developments in computer technology and
prograrnrning, ii was agreed to limit Lhe scope of lhe new list to dcfinitions, explanatory coni-
rnentary, and iliustrations of wood anatomical descriptors, rather Lhan concentrate on codiiig
procedures.
A new Committee was appointed by lhe IAWA Council to work towards Lhe new list, and
thanks to a substaruial grant from lhe USDA Competidve Research Granis - Wood Utilization
Program (Grant No. 88-33541-4081), a workshop was held by lhe Committee from October
2-7, 1988, in lhe Department of Wood & Paper Science, North Carolina State University,
Raleigh, NC. USA, under lhe joint auspices of IAWA and IUFRO Division 5. A preiiminary
list was prepared during lhe workshop. IAWA rnembcrs were invited to cornment on this list,
and these comrnents helped with lhe final preparation of Lhe new li st. Pie hst presented here was
agreed to after review of subsequent drafts and extensive internal consuitation between conimo-
tee members.
Although ihis list has 163 anatomical and 58 misceilaneous features, it is not a complete li,[
encompassing ali Lhe structural patterns that one can encounter in hardwoods. lnstead it is in-
tended to be a concise list of features usefui for identification purposes. Also, lhe numbers as
signed to each feature in lhe present list are not meant to be codes for a computer program, but
are intended to serve for easy reference, and to help translate data from one program/database tu
another.
Wood and wood celis are biological elements, formed in trees, shrubs, and climbcrs to fui! dl
a physiological or niechanicai function. Aithough there is more discrete diversity in wood struc-
ture than in many other plant parts, there is also much continuous variation, and any auempt to
ciassify this diversity into well-defined features has an artificial eleinent. Yet we are confident
that in Lhe feature list presented here ambiguity of descriptors has heen iimited to a minimiitn.
and we hope that ali present and future coileagues engaged in wood and decripti\ e
wood anatomy will find this list a valuable guide and reference.

The IAWA Committee:

VERONICA ANGYALOSSY ALFONSO
Divisão de Madeiras, I.P.T. Cidade Untversitaria, Sio Paulo. Brazil

PIETER BAAS
Rijksherbariuni, Leiden, 'lhe Netherlands
SHERWIN CARLQUIST
Rancho Santa Ana Bo:anic Garden, Clarculont, Ca!ifornia, U.S. A.

1 AWA List of microscopíc features for hardwood identification 223
222 IAWA Bulle.tin n..%., Vol. 10 (3), 1989

ACKNOWLEDGEMENTS
JOAO PERES CHIMELO
Divisão de Madeiras, I.P.T. Cidade Universitária, São Paulo, Brazil
The IAWA Cornrnittee is greatly indebted to the following institutions and individuais:
VERA T. RAUBER CORADIN
Instituto Brasiliero de Desenvolvimento Florestal, Departmento de Pesquisa, Brasilia, Brazil The USDA Competitive Rescarch Grants-Wood Utilization Program (Grant No. 88-33541-
4081) for financing the IAWA/1UFRO Workshop in Raleigh, North Carolina, and subsequent
PIERRE DÉTIENNE
rneetings in London and Leiden by P. Baas, P.E. Gasson, and E. A. Wheeler.
Division d'Anaomie des Bois, Centre Technique Forestier Tropical, Nogent-sur-Mame, France
The Department of Wood and Paper Science, N. C. State University for offering hospitaiity
PETER E. GASSON
and facilities during the IAWA/IUFRO Workshop in Raleigh; especially Dr. C.A. LaPasha and
Jodreil Laboratory, Royai Botanic Gardens, Kew, U. K.
Ms. Vann Moore for help with preparation of the various drafts, and Ms. Milie Sullivan.
DIETGER GROSSER
The Forest Products Laboratory, Madison, Wisçonsin, USA, for providing financial support
Institui fiir Holzforschung und Holztechnik der Universitiit München, Miinchen, F.R.G.
towards the printing costs of this special issue,
JUGO ILIC
The Jodreli Laboratory, Royai Botanical Gardens Kew, UK, for supporting photographic
CSIRO, Wood Scicnce & Technology, Highett, Victoria, Australia
work, and providing facilities and hospitality during a mecting in March 1989 for the selection
KEIKO KURODA of iliustrations.
Forestry & Forest Products Research Institute, Kansai Branch, Kyoto, Japan
The BaiIey-Wctmore Laboratory of Plant Anatomy and Morphology, harvard University,
REGIS B. MILLER and Dr. P. B. Tomhinson, Dr. D. Pfister, and Dr. A. Knoll for giving access to the Bailey nega-
Center for Wood Anatomy Research, Forest Products Laboratory, Madison,Wisconsjn, U.S.A. tives and darkroom facihities.
KEN OGATA The Rijksherbarium for various facilities; especially to Ms. Emma E. van Nieuwkoop for
Wood Technology Division, Forestry & Forest Products Research Institute, Tsukuha, Japan mounting the plates, and lay-out editing.
HANS GE0RG RICHTER All IAWA Members who have kindly given their comrnents on various drafts of this list:
Institui für Holzbiologie und Holzschutz, Bundesforschungsanstalt für Forst- und Holzwirt-
schaft, Hamburg, F.R.G. K.M. Bhat, India Yvonne Hemberger, Harnburg, FRG
Lim Seng Choon, Kepong, Malaysia Alberta M.W. Mennega, Utrecht, The Netherlands
BEN J. H. TER WELLE
D. F. Cutler, Kew, UK C.A. LaPasha, Raleigh, NC, USA
Rijksuniversiteit Utrecht, Instiruut voor Systeniatische Plantkunde, Utrecht, Thc Netheriands A. Londono, Columbia
W, C. Dickison, Chapel Hill, NC, USA
ELISABETH A. WHEELER T. Fujii, Tsukuba, Japan Paula Rudail, Kew, UK
Department of Wood & Paper Science, North Carolina State University, Raleigh, North Caro- H. Gottwald, Hamburg, FRG M. Seth, india
una, U.S.A. Mary Gregory, Kew, UK

AcknowuedgemefltS for iflustrations

Photograplis by courtesy of:

I. W. Bailey, Bailey -Wetmore Laboratory of Plant Anatomy and Morphoiogy, Harvard Uni-
versity: 10, 11, 16, 18, 39, 57, 58, 64, 65, 148.
Blumea: 38, 44, 73, 74 (Baas 1973), 174 (Van Vliet 1981).
P. Détienne: 129.
P.E. Gasson: 2,4,7, 8, 12, 19, 21, 26, 28, 30-34, 36, 37, 40, 45-54, 63, 66, 75, 78-82,
84-86, 88, 90-93, 95-99, 102-106, III, 114-116, 118, 120, 122, 126-128, 130-
135, 137-144, 151, 153, 154, 156, 157, 159, 161, 163-168, 171, 172, 176, 178, 180-
182, 188.
D. Grosser: 15, 27, 29, 55, 68, 71, 72, 112, 113, 146, 158, 170, 173, 177.
 225
TAWA Builetin nt., Vol. 10 (3), 1989 1 AWA List of microscopie features for hardwoodidefltiflCatiOfl
224

1 AWA Buileun: 3 (Bricigwater & Baas 1982), 35 (Vidal Comes ex ai. 1988), 70 & 123 (Bridg
water & Baas 1982), 155 (Topper & Koek-Noorman 1980), 175 Baas et ai. 1988), 184
(Gottwald 1983), 185 (Ter WelIe 1980).
J. lhe: 56.
C.A.LaPasha: 190.
R. B. Milier: 160, 186, 187, 189.
K. Ogata: 1, 5, 9, 13, 14, 20, 22, 24, 25, 41, 42, 61, 62, 76, 77, 83, 89, 94, 101, 107-109,
117, 119, 124, 125, 136, 145, 147, 149, 150, 152, 162, 179.
E. A. Wheeler: 6, 17, 23, 43, 59, 60, 67, 87, 100, 110, 121, 169, 183.
H. P. Wilkinson: 69.
EXI'LANATORY NOTES
Quantitative FeatureS - For quantitative features of general applicability (e.g., vessel
frcquency, tangential vessel lumen diameter, vessel element hength, and fibre length), this list
ineludes broad categories for easy use when identifying unknowns, as well as more precise
quantitative descriptors (mean, range, standard deviation). When constnicting a database the
nutnbers of samples as well as Lhe number of measurements or counis done per samplc should
be recorded Different computer programs allow storage of different amounts of information
(eg., ali measurements, orjust the means, ranges, and standard deviations), and use different
algorithms for rnatching quantitative features. This publication does not recommend a particular
program or a particular method for the storage and retrieval of quantitative data, but provides
some guidance on how to obtain these data.

Variable Features and Relative Abundance - l3ecausc of wood's inherent variabihi-

ty, it is inevitable that some fe.atures will be well-defined in some samples while absent or iii-
defined in other samples of the sarne species. Accommodatiflg such variability has always beco a
problem in key constiUctiofl, and most keys (computerised or otherwise) have provisions for
such situations. Describing relative abundance of some features, e.g., prismatic crystals, is also
problematic, and textual cominents on reiative ftequency should be added to a description or
database- In this list of descriptors, some features apply only when the characteristie is of com-
moo occurrence. For such features, the illustrations and examples are intended to help interpret
'common'. Although many keys have used these sarne features accompanied by the sarne quali-
fier 'common', there have been no extensive systernatic analyses to determine what per cent
ocçurrence constitutes 'common'. Therefore no quantitative criteria for 'common' have been
offered in this list.

IAWA Bulietin n,s., Vol. lO (3), 1989 IAWA.Listof microscopie features for hardwood idcntification 227
226

UST OF FEATURES Vessei-ray pirring - p. 253

Name 30. Vessel-ray pits with distinci borders; similar to intervessel pits in size and shape throughout
Lhe ray celi
ANATOMICAL FEATI:RES 31. Vessel-ray pits with much reduced borders to apparently siniple: pits rounded or angular
32. Vessel-ray pits with much reduced borders to apparently simple: pits horizontal (scalari-
Growth rings - p.234 forrn, gash-like) to vertical (palisade)
33. Vessel-ray pits of two distirict sizes or types in Lhe sarne ray cdl
1. Growrh ring boundaries distinct 34. Vesscl-ray pus unilaterally con'ipound and coarse (over 10 um)
2. Growth ring boundaries indisiinct or absent 35. Vessel-ray pits restricted to marginal rows

Vesseis - p. 236 Helica! thickenings - p. 256

36. Ilelica] t1'iickenings in vessel elernents present
Porosity p. 236 37. Helical thickenings throughout body of vesse] elernent
3 Wood ring-poroiis 38. Helical thickenings only in vessel element tails
4. Wood semi-ring-porous 39. Helical thickenings only in narrower vessel elernei:ts
5. Wood diffuse-porous
Tangential diameter of vessel lwnina p. 25
Vessel arrangeinent - p. 238
Mean tangential diameter of vessel luniina
6. Vesseis in tangential hands 40. 2^50um
7. Vessels in diagonal and/or radial partem 41. 50-100nn
8. Vesseis in dendritic pattern 42. 100-200 .riii
Vessel groupings - p. 242 43. ^200p.m
9. Vesseis exclusively solitary (90% or more) 44. Mean, +1- Standard Deviation, Range, r - x
10. Vessels iri radial multiples of4or more common 45. Vesseis of mo distinct diameter classes, woçd no: nng-pero:s
11. Vessel clusters common
Vesseis per square rnilürnetre - p. 259
Solitary vessel outiine - p. 244 46, 5 5 vesseis per square millinieue
12. Solitary vessel outline angular 47. 5-20 vessels per square rnillimetrc
48. 20-40 vessels per square millirnetre
Perforation plates p. 246
49, 40-100 vesseis per square rnillimetre
13. Simple perforatiori plates
14, Scalariform perforation plates 50. ^ 100 vcssels per square rnillimetre
51, Mean, +1- Standard Deviation, Range, n = x
15. Scalariform perfor'ation plates with ^ 10 bars
16. Scalariforrn perforation plateswith 10-20bars Mean versei elernent iengrh - p. 259
17. Scalariform perforarion plates with 20-40bars 52. ^350p.m
18. Scalarit'onn perforation piates with ^ 40 bars 53. 350-800 piri
19. Reticulate, foraminate, and/or other types of multiple peiforatiori plaes 54. ^800sm
lnteri.-essel pus: arrangernent and sire - p. 250 55, Mcan, +/- Standard Deviauon, Range. n = x
20. Intervessel pits scalariforrn
21. Inrervessel pus opposite
Tyloses and deposit.s ia vesseis -p. 259
22. Intervessel pits alternate 56. Tyloses common
57. Tyloses sclerotic
23. Shape of alternate pits polygonal
24, Minute - 58. Gums and other deposits in heartwoxl vessels
25. SmaU - 4-7 rn Wood vesselless - p. 262
26. Medium - 7-10prn 59. Wood vesselless
27. Lrge - ^:lOi.tm
28. Range of iniervessel pit size (m) Tracheicis and fibres - p262
Vestured pits '- p. 252
29. Vestured pits 60. Vascular/vasicenflic tracheids present

228 1\\vj\ OLilietin n.s., l(.) (i, 989 \\V.\ [_isL of nicroscpic caui tor Iidood

Ground tissue fibres - p. 264 Rays -p.282

61. Fibres with simple to minutely bordered pits
62. Fibres with clistinctly bordered pits Ray width - p. 282
63. Fibre pits common in both radial and tangential walis 96. Rays exciusively uniseriate
64. Helical thickenings m ground tissue fibres 97. Ray width 1 to 3 cells
98. Largerrays commonly 4- to lO-seriate
Septate fibres and parenchyma-like fibre band - p. 266
99. Larger rays comrnonly> lO-seriate
65. Septate fibres present
100. Rays with multiseriate portion(s) as wide as uniseriate portions
66. Non-septate fibres present
67. Parenchyma-like fibre bands alternating with ordinary flbres Aggregare rays - p. 284
101. Aggregate rays
Fibre wall rhickness - p. 268
68. Fibres very thin-walled Ray height p. 284
69. Fibres thin- to thick-walled 102. Ray height> 1 mm
70. Fibres vel'y thick-wallcd
Rays of iwo distinci sizes - p. 286
Mean fibre lengths - p. 269 103. Rays of mo distinct sizes
71. 5900i.tm
72. 900-1600 jim Rays: celiular co,nposirion - p. 288
73. >1600j.tm 104. Ali ray celis procumbent
74. Mean, +/- Standard Deviation, Range, n = x 105. Ali ray celis upright and/or square
106. Body ray celis procumbent with one row of upright and/or squarc marginal celis
Axial parenchyma - p. 270 107. Body ray cells procumbent with mostly 2-4 rows of upright and/or square marginal cells
108. Body ray edis procunibent with over 4 rows of upright and/or squarc marginal celis
75. Axial parenchyma absent or extremely rare 109. Rays with procumbent, squarc and upright cclls mixed throughout the ray

Apotracheal axial parenchynia - p. 270 Shearh celis - p. 292

76. Axial parenchyrna diffuse 110. Sheath celis
77. Axial parenchyma diffuse-in-aggregates
Tile celLs - p. 292
Pararracheal axial parenchyma - p 272 111. Tile cells
78. Axial parenchyma scanty paratracheai
Pe,forared ray cells - p. 294
79. Axial parenchyma vasicentric
112. Perforated ray edis
80 Axial parenchyma aliform
81. Axial parenchyma lozenge-aliform Disjunctive ray parerzchvma ccli wall.ç - p. 294
82. Axial parenchyma winged-aliform 113. Disjunctive ray parenchyrna ceil walls
83. Axial parenchyma confluent Rays per ,nillimetre - p. 296
84. Axial parenchyma unilateral paratracheal 114, 54/mm
Banded parenchyma - p. 276 115. 4-12/mm
85. Axial parenchyma bands more than three celis wide 116. ^ 121mm
86. Axial parenchyma in narrow bands orlines up to three cells wide
Wood rayless - p. 297
87. Axial paxenchyma reticulate
117. Woodrayless
88. Axial parenchyma scalariform
89. Axial parenchyma in marginal or in seemingly marginal bands
Storied structure - p. 298
Axial parenchyma ceil type/srrand length p. 280
90. Fusiform parenchyma cells 118. All rays storied
91. Two celis per parenchyma sirand 119. Low rays storied, high rays non-storied.
92. Four (3-4) cdlis per parenchyma strand 120. Axial parenchyma and/or vessel elements sto:
93. Eight (5-8) cells per parenchyma strand 121. Fibres storied
94. Over eight cdlls per parenchyma sirand 122. Rays and/or axial elements irregularly storie
95. Unlignified parenchyma 123. Number of ray tiers per axial mm
 1\V,\ Bulk'tin r., 'Vol. li) ). 1)') 1 .\\V1\ List ol nhicroscopic icatures fnr i irdocd dcnti:cr

Secretory elements and cambial variants -p300 Silica P. 318

159. Silica bodies preseni
OU and mucilage celis - p. 300 160. Silica bodies in rav colis
124. Oil and/or mucilage cdlis associated with ray parenchyma 161. Silica bodies in axial j cnchvnm c cis
125. 011 andj or rnucilage cells associated with axial parenchyrna 162. Silica boclies iii fibres
126. Oil and/or mucilage cells presem among fibres 163. Vitreous silica
Inrerceiluiar canais - p. 302
127. Axial canais in long tangential limes APPENDIX - Non-anatornical information p. 2 1
128, Axial canais in short tangential lines
129. Axial canais cliffuse Geographical distribution - p. 321
130. Radial canais 164. Europe and temperate Asia (Brazier and Franklin region 74)
131. Intercellular canais of traumatic origin 165. Europe, excluding Mediterranean
Tubes / tubuies - p. 306 166. Mediterranean including Northern Africa and Middle East
132. Laticifers or tanniniferous tubes 167, Temperate Asia (China), Japan, IJSSR
168. Central South Asia (Brazier and Franklin region 7)
Cambial varians - p. 308 169. India, Pakistan, Sri Lanka
133. Included phloem, concentric 170. Bumaa
134. Included phloem, diffuse
171. Southeast Asia and the Pacific (Brazier and Franklin region 76)
135. Other cambial variants
172. Thailand, Laos, Vietnam, Cambodja (Indochina)
173. Indomalcsia: Indonesia, Philippines, Malaysia, Brunei, Papua New Guinca, and
Mineral inclusions - p. 310
Solomon Jslands
174. Pacific Islands (including New Caledonia, Samoa, Hawaii. and Fiji)
Pris,naric crystals -p. 310
175. Ausu'alia and New Zcaland (Bra7.ier and Franklin region 77
136. Prisrnatic crystals present
137. Prisrnatic crystals in upright and/or square ray crus 176. Australia
138. Prismatic crystals in procumbcnt ray celis 177. NewZealand
139. Prismatic crystals in radial alignment in procumbent ray cells 178. Tropical mainland Africa and adjaceni isla,ids (Brazier and Franklin region 78)
140. Prismatic erystais in chambered upnght and/or square ray celis 179. Tropical Africa
141. Prismatic crystals in non-charnbered axial parenchyma cells 180. Madagascar & Mauritius, Runion & Cornores
142. Prismaiic crystals in chanibered axial parenchymacells 181. Southem Africa (south of the Tropic of Capricorn) (BraLier and 1-ranklin region 70)
143. Prismatic crystals in fibres 182. North America, north of Mexico (Braaler and Franklin region $0)
183. Neotropics and temperate Brazil (Brazier and Franklin rcgion 8 II
Druses - p. 313
184. Mexico and Central America
144 Druses present
145. Druses in ray parenchyrna celis 185. Carrubbean
146. Druses in axial parenchyrna cells 186. Tropical South America
147. Druses in fibres 187. Southern Brazii
148. Druses in chambered cells 188. Temperate South America including Argentina. Chile, Uruguav. and S. Paraguav Brazicr
and Franklin region 82)
Olher crystai types -- p. 313
149. Raphides Habit - p. 321
150. Acicular crystals 189. Tree
151. Styloids and/or elongate crystals 190. Shrub
152. Crystals of other shapes (mostly small) 191. Vine/liana
153. Crystal sand
Wood of co,nmerciai importance - p. 322
Orher dia gnoszic crystal features - p. 315 192. Wood of commercial irnportance
154. More than one crystal of about the sarne size per celi or chamber
155. Two distinct sizes of crystals perceil or chamber Specfl c gravity . p. 322
156. Crystals in enlarged cells 193. Basic spccific gravity iow, S 0.40
157. Ciystals in tyloses 194. Basic specific gravity medium, 0.40-0,75
158. Cystoliths 195. Basic spccific gravity high, ^ 0.75
 jmw-

232 IAWA Bulletin n. s. Vol. 10 3), 1989 1 A\VA. List nf rnicroscopic featurts for hardwocd identification 23

Hearwood colour - p. 323

196. Heartwood colour darker than sapwood colour NAME
197. Heartwood basically brown or shades of brown
198. Heartwood basicaily redor shades ofred Family, genus, species, authority.
199. Heartwood basically yeilow or shades of yeliow
200. Heartwood basically white to grey When creating a database iL is essential to record the fuli taxonomic information on the speci.
mens, i.e., record farni!y, genus, species, authority. For authorities foilow curnmonly used
201. Heartwood with streaks
202. Heartwood not as above abbreviations (listed in Mabberley 1987). Reference to Willis's Dictionary of Fiowering P!ants
and Fems(Willis 1973) and Mabberley (1987) is helpful in determining the farnilia) affinities of
Odour - p. 325 various genera, and preferred fami!y names. When preparing a database, a!so indicate which
203. Distinct odour particular classification scheme, with respect to fami!y delimitation, is heing used (e. o., Takhta-
Heartivood fluorescence - p. 325 jan 1980, 1987; Cronquist 1981, 1988; Thorne 1976).
204. Heartwood fluorescent li can be usefui to retrieve information on the wood anatomy of particular fami!ies or to re-
strict the search for the identity of an unknown wood to a certain farni!y or farni!ies. Conse-
Warer & ethanol extracrs: fluorescence & colour - p. 326 quently, it is advisable to record the family as a feature. For the family name it is not critical
205. Water extract fluorescent what method of coding is employed so long as it is cleariy explained in notes acconlpanying the
206. Water cxtract basically colour!ess to brown or shades of brown database. Por instance, the family can he indicated as 3-1etter acron yms (Weber 1982) or as
207. Water extract basically redor shades ofred
numerical codes (see pp. 127 and 144-145 in Milier 1981).
208. Water extract basically yeilow or shades of yellow
209. Water exuact not as above
210. Ethanol extract fluorescent
211. Ethanol extract basically colour!ess to brown or shades of browri
212. EthanoÏ extract basically red or shades of red
213. Ethanol extract basically yellow or shades of yellow
214. Ethanoi extract not as above
Frorh rest —p.327
215. Froth test positive
Chrorne Azurol-S test - p. 328
216. Chrome Azurol-S test positive
Burning splinter test - p. 328
217. Splinter burns to charcoal
218. Splinter burns to a fuli ash: Colour of ash bright white
219. Spiinter burns to a fui! ash: Colour of ash yellow-brown
220. Splinter bums to a fuli ash: Colour of ash other than above
221. Splinter burns to a partial ash

234 1 AWA List of microscopic fé-atures for hardwood identification 235
IAWA Builetin lis., Vol. 10(3). 199

ANATOMICAL FEATURES

GROWTH RTNGS

1. Growth ring boundaries distinct

2. Growth ring boundaries iridistinct or absent

Deflnitions:
Growth ring boundaries distinct = growth rings with an abrupt structural change at thc
boundaries between them, usually including a change in fibre wall thickness and/or fibre radial
diameter, figs. 1, 2.
Growth ring boundaries indistinct or absent = growth rings vague and marked by
more or less gradual structural changes ai their poorly defined boundarics, or not visible, fig. 3.

Comments:
Growth ring boundanes can be marked by one ar more of the following structural changes:
a. Thick-walled and radially llattened latewood fibres ar tracheids versus thin-walled early-
wood fibres ar tracheids, fig. 1, e. g., Weinmannia trichosperma (Cunoriiaçeae), Laurus no-
bílis (Lauraceae).
b. Marked differcnces in vesse] diarneter between latewood and earlywood of the following
ring as in semi-riiig-porous and ring-parous woods, figs. 5-8, e.g., Juglans regia (Jug]an-
daceae), Ulmus procera (til niaceae).
c. Marginal parenchyma (terminal or initial), fig. 2, e.g.,Xylopia nítida ( Annonaceae), Bra-
chystegia Iauren:ii (Caesalpiniaceae), Juglans regia (Juglandaceae), Liriodendron tuliptfera
(Magnoliaceae). lrregularly zonate, tangential parenchyma bands witfioui associated abrupi
changes in fibre diameter or wall thickness are not considered marginal and do not represeni
discinct growth ring boundaries, e.g., Eschwei lera subglandulosa (Lecythidaceae), Irvingia
excelsa (S imaroubaeeae).
d. Vascular tracheids and very narrow vessel elements very numerous or forming the ground
tissue of the larewood, and absent from the earlywood, e.g., Sanbucus nigra (Caprifolia.
ceae).
. Decreasing frequency of parenchynia bands towards the latewood resulting in distinct fibre
zones, e. g.. Lecythis pisonis (Lecythidaceae), Doneila pruniformis (Sapotaceae).
t. Distended rays, e. g., Fagus spp. (Fagaceae).
See Carlquist (1980, 1988) for other types of growth ring boundaries and for commonly
occurring combinations of several of the above fearures.
.Although absence of growth ring boundaries is a clear enough descriptor, the differences be-
en 'indistinct' and 'distinct' boundaries are somewhat arbitrary, and there are intermediates
:g. 4). Growth nngs may appear distinct when observed macroscopically, yet have indistinci
boundaries ai the light microscopic leveI; distinctness of the ring boundaries should be judged
with a microscope. Indistinct growth ring boundaries are very cornmon in tropical trees (fig. 3. Figs. 1 & 2. Growth ring boundaries distinct (feacure 1). - 1: Weinmunnia trichosperma, boun-
g., Spondias mombin - Anacardiaceae, Parkia nítida -- Min,osaceae, Coelocarvon preuasii dary niarked by differences in fibre and vessel diniensions, x 80. - 2: Xylopia nítida, boundary
\tvristicaceae; Xanrhophvllum philippinen.ve - Polygalaceae). rnarked by thick-wallcd latewood fibres and marginal parenchyma band, x 48. - Fig. 3.
Nonperiodical, sporadic occurrence of ring boundaries (due to unusual clitnatic extrenies or Growib ring boundaries indistinct or absent (feanire 2), Xanthophyllum philippinense x 22. -
niaic cvcrlts) should he recorded as rings ahsent or bouridaries indist inct. Fig. 4. Growch ring boundaries intermediate between distinct and indisunct (features 1 and 2
variable), Jacaranda copaia, x 48.
 236 IAWA Bulletjn fl.S., VOL 10 (3), 1989
IAWA List of microscop jcfeatures for hard ,o idcntificatjon
237

POROSITY

3. Wood ring-porous
4. Wood semi-ring-porous
5. Wood diffuse-porous

Definizions:

Wood ring-porous = wood in which the vesseis in the earlywood are distinctly larger
than those in the Iatewood of the prcvious and of the sarne growth ring, and foi-rn a well defined
zone or ring, and in which there is an abrupt transition to the latewood of the sarne growth ring,
fig. 5, e. g., Quercus robur (Fagaceae), Frw(nus exce(sjor
(Oleaceae), Phellodendron a,nurense
(Rutaceae), Buineija lanuginosa (Sapotaceae), Ul,nu.s americana
(Ulmaceae).
Wood semi-ring-porous = 1) wood in which the vesseis in the earlywood are distinctly
larger than those in the Iatewood of the previous growth ring, but in which there is a gradual
change to narrower vesseis in Lhe intermediate and latewood of the sarne growth ring; or
2) wood with a distinct ring of closely spaced earlywood vesscls that are not markedly larger
than the Iatewood vessels of the preceding ring or the sarne growth ring. Alternative definition:
intermediate condition between ring-porous and diffuse-porous wood, figs. 6, 7, e.g., Cordia
trichozona (Boraginaceae), Juglans nigra (J ugi andaceae).
Lagersrrocmialf (Lythra-
ceae), Cedrela odorara (Meliaceae), Prerocarpus indicas ( Papilionaceae), oribunda
Prunus amygdalus
(Rosaceae), Paulownja romentosa (Scrophujarjaceae)

Wood diffuse-porous = wood in which the vesseis have more or less the sarne diarneter
throughout the growth ring, figs. 9, 10, e. g., Acer spp. (Aceraceae), RJ2ododefldron wadanum
(Encaceae), CercidipJiyllu,njapjcum (Cercidiphyllaceae),
Swjetenia spp. (Meliaceac), Entero.
lobium spp. (Mirnosaceae); the vast rnajority of tropical species and most temperate species.

Comments:
The three features for porosity form an intergrading continuum and many species range from
diffuse-porous to semi-ring . porous, or from ring-porous to semi-ring-porous. Porosity (fea-
iures 3-5) is coded independently of vessel arrangenient (features 6-8).This iniplies that
woods with a distinct vessel arrangernent (features 6-8), as well as those with evenly dis-
Uibuted vesseis, may be diffuse-porous.
In some temperate diffuse-porous woods (e- g, Pagas spp. —Fagaceae, Platanu.s spp. -
Plalanaceae) the Iatest formed vesse]s in the latewood may be considerahly smaller than those of
the earlywood of the next ring, but vessel diarneter is more or less uniforrn throughout rnost of
the growth ring (fig. 10).
In a description, characteristjcs of the earlywood ring of ring-porous woods should be
noted, i.e., describe how rnany vessels wide the ring is. Sudo's (1959) key used the features
'pore ring: l-seriate' and 'pore ring: multiseriate'. Such characteristics can be useful in distin-
guishing between species, e.g., Ulmus americana Fig. 5. Wood ring-porous (feature 3), Phellodendron amurense, x
typically has an earlywood zone that is one 28. - Figs. 6 & 7. Wood
vessel deep, Ulrnus rubra has ai earlywood zone that is more than two vesseis deep. semi-ring-poroti s (feature 4). —6: Prunus sp., x
25. - 7: Paulownia tomenzosa, x 18. - Fig. 8.
Wood ring-porous (feature 3), bui with narrow rin g s givino false irnprcssion of di ffuc- nr
Caurion: Slow grown ring-porous woods have narrow growth rings with very little latewood semi_nngporsj. Caralpa binon [odes.
x 30.
(fig. 8). Be careful not to confuse the closely spaced earlywood zones of slow-grown ring-
porous woods with a tangential partem, or to interprel such woods as diffuse-porous.

--

IAWA List of microscopic features for hardwooci identification 239
238 IAWA Builetiri n. s., Vol. 10(3),_1989

Figs. 9 & 10. Wood diffuse-porous (feature 5), - 9: Rhododendron wadanum, x 75. - 10:
Cercidiphyllwn japonicum, classified as diffusc-porous despite di fference s in vcssel diarneter of
Iatest forrned latewood and earlywood. x 30.
'.4
VESSEL ARRANGEMENT

6. Vesseis in tangential bands

7. Vesseis in diagonal and for radial pattern
8. Vesseis in dendritic pattern

Definirions:
Vesseis in tangential bands = vesseis arranged perpendicular to the rays and forrning
shori or lon og tangential bands; these bands can be straight or wavy; includes ulrniforrn and fes-
tooned, figs. 11-13, e.g., Kalopanax piclus (Araliaceae), Patagonulu americana (Boragina-
ceae), Enkianthus cornuus (Ericaceae), Madura pom (fera (Moraceae), Pirtosporusn tobira (Pino-
sporaceae), Cardwellia sublirnis (Proteaceae).
Vesseis in diagonal and/or radial pattern = vesseis arranged radialiy or intermediate
between tangential and radial (i. e., oblique), figs. 14, 17, 20. e. g., Lithocarpus edulis (Faga-
ceae), Calophyllum brasiliense, C. papuanwn, Mesuaferrea (Guttiferae), Eucalyptus diversi-
colar, E. obliqua (Myrtaceae), Amyris sylvatica (Rutaceae), Chloraluma gonocarpa (Sapota-
ceae). Synonym for diagonal: in echelon'. Figs. 11-13. Vesseis in tangential bands (feature 6). —11: Latewood vesseis in tangential bands
Vesseis in dendritic pattern = vesseis arranged in a branching patiem, forining distinct (note aiso feature 3, wood ring-porous), Kalopanax picrus, x 80. - 12: Vesseis and parenchyma
'festooned', Cardwellja sublimis, x 30, - 13: Ali vesseis in tangential bands, Enkianrhus
tracEs, separated hy arcas devoid of vesseis, figs. 15, 16, e. g, Rhus aromarica (Anacardiaceae),
cornuus, x 75. Fig. 14. Vessels in a diagonal patterri (feature 7), Calophyllum papuanuin,
Castanea dentara (Fagaceae), Chionanihus retusus (Oleaceae), Rhamnu.s carhartica (Rhamna- x 29.
ceae), Bumelia lanuginosa (Sapotaceae). Synonym: ti ame-like.
240 IAWA Builetin n.s., Vol. 10(3), 1989 IAWA List of microscopic features for hardwood identification 241

• . . .. O

Fig. 19. Vesseis in a diagonal to dcndritic pattcrn (fcatures 7 and 8), Burnelia ohwifo1ia (n(Ite
also featiire 5, wood diífuse-porous), x 45. - Fig. 20. Vesseis in a diagonal tu radial pattcrn
(feature 7), Lirhocarpus edulis, x 29.

. vessel distribtnion pattems.
:•.
k ti &q4.
a '1
18
Figs. 15 & 16. Vesseis in a dendritic pattern (feature 8). - 15: Rhamnus cathartjca (note also
feature 5, wood diffuse-porous), x 60. - 16: Rhus arornatica, clendritic pauern restrictcd to
Iatewood vesseis (note also feature 3, wood ring-porous), x 35. --- Fig. 17. Vessels ir] a
radial pattern (feature 7), Amyris sy/vazica, x 18. - Fig. 19. Narrow vesseis in a tangential io
diagonal pattern (features 6 and 7), Kalopanax picrus (note also feature 3, wood ring-porous),
x 80.
Procedure:
Vessel disnibution patterns (tangential, diagonal /radial, dendritic) are determined from the
cross section at a low magnilication, and are recorded only where there is a distinct pattern. In
ring-porous woods, only the internied iate- wood and latewood are earnined. The ring ofvessels
at the beginning of lhe growth ring of ring-porous woods is not considered when determining
Commenrs:
These features often occur iri cornbination, Vesse] arrangement in some woods intergrades
between tangential and diagonal (fig. 18). Diagonal and dendriric oftcn iritergrade (fig. 19). AlI
applilable features shou]d be recorded.
The arrangement of pore clusters seen in most species of Ulmus (Ulmaceae) lias been called
ulmiform; thjs describes woods where the latewood clusters are predorninantly in wavy tan-
gential hands (feature 6) and sornetimes temi to a diagonal patern (feature 7). Tangential ares oí
vesseis, typical of the Proteaceae (fig. 12), have heen cal]ed festooned.
Since, in ring-porous temperate spccics, these patterns (features 6-8) may be resuicied tu
the latcwood, their expression depends on ring width, and when rings are narrow these patterns
are not obvious.

242 IAWA Builetin n,s., Vol. 10 (3), 1989 IAWA List of microscopic features for hardwood identification 243

VESSEL GROUPINGS

9. Vessels exelusively solitary (90% or more)

10. Vesseis in radial multiples of 4 or more common
11. Vessel clusters common

DeJinitions:
Vessels exclusively solitary = 90% or more of the vessels are completcly surrounded
by other elernents, i. e., 90% or more appear not to contact another vessei, as viewed in cross
section, fig. 21, e. g., Aspidosperina quebracho (Apocynaceae), Caraipa spp. (Bonnctiaceae),
Eucalyptus regnans (Myrtaceae), Malus sylvestris (Rosaceae), Schirna wallichii (Theaceae).
Radial multiples o! 4 or more cornmon = radial files of 4 or more adjacent vesseis of
common occurrence, fig. 22, e.g., Cerberafloribunda (Apocynaceae), Ilex aquifõliwn (Aqui-
foi iaceae), Brachylaena hutchin.sii (Compositae), Elaeocarpus hookerianus (Elaeocarpaceae),
Strychnos nux-vomica (Loganiaceae), Arnyris balsamjfera (Rutaceae), Gambeya excelsa (Sapo-
taceae).
Clusters common = groups of 3 or more vesseis having both radial and tangential con-
tacts, and of common occurrence, fig. 23, e.g., Polyscias elegans (Araliaceae), Pirtosporurn
ferrugineum (Pittosporaceae), latewood of Glediisia triacanihos, Gymnocladus dioica (Caesai-
piniaceae), Morus alba (Moraceae), and Ailanrhus altissima (Sirnaroubaceae).

Com.'nenzs.'
Feature 10 'radial multiples of 4 or more common' should be used only when radial muitiples
of 4 or more are an obvious feature of the transverse section. Feature 11 'clusters common' ap-
plies only when clusters are frequent enough that they are easily observed during a quick scan of
a cross section. Clusters and radial rnultiples of 4 or more are not rnutually exclusive and can
occur in combination. Woods with vesseis in tangential bands (feature 6) ofren have clusters.
The most common vessel grouping is radial multiples of 2 to 4 wirh a variable proportion of
solitary vessels (fig. 24). The absence of features 9-11 automatically irnplies this condition,
When describing a wood, ao index of vessel grouping can be calcuiated in the manner re-
comrnended by Carlquist (1988): count the total number of vesseis in a minimum of 25 vessel
'groups' (i.e., count both solitary vesseis and vessel multiples as a 'group'), divide the total
number of vesseis by 25 (the number of groups counted). An index of 1.00 indicates exclusive-
ly solitary vessels, and the higher the index, the greater the degree of vessel grouping.

Caution: Care is needed to recognise rhe foliowing as not being multiples: (i) solitary vesseis Fig. 21. Vesseis 'cxclusivcly solitary' (feature 9), Aspidosperrrza quebracho, x 45. - Fig. 22.
composed of vessel elements with oblique overlapping end walis giving the appearance ofvessel Radial rnultiples of 4 or more cominon (feature 10), Elaeocurpus hookerianus, x 29. Fig. 23.
pairs on the cross section as in Cercidiphyllurn (Cercidiphyllaceae) and Illicium (llhiciaceae), and Clusters common (feature 11), Gymnocladus dioica, latewood, x 140. - Fig. 24. Vesseis part-
(ii) closely associated solitary vessels, as in some species of Eucalyptus (Myrtaceae) and Calo- ly solitary, partiy in radial multiples of 2-4, or very small c1usters (features 9, 10, and 11
phyllurn (Guttiferae) (Brazier & Franklin 1961). absens), Drypetes gerrardii, x 75.

244 IAWA Bu]letin os., Vol. 10 (3)1989 IAWA List of microscopic features for hardwood identification 245

SOL1TARY VESSEL OUTLINE

12. Solitary vessel outline angular

Definition.
Solitary vessel outline angular = shape oU so]iary vessel outline is angular as viewed
in cross section, fig.25, e.g., Aexroxkonpuncratum (Aextoxicaceae), Cercidiphyllwnjaponi-
cum (Cercidiphyllaceae). Iatewood vesseis of white oaks (e. g., Quercus alba, Q. robur—Faga-
ceae), Stemonurus luzoniensis (Icacinaceae), Hortonia spp. and Mol1ini spp. (Monimiaceae).

Procedure:
In ring-porous woods, examine the latewood because in these woods the earlywood vesseis
are almost always circular to oval in outlije. Use Lhe outline of the solitary vesseis because the
cornrnon walis of vesseis in inultiples can be flattened giving pari oU the vesseis ao angular oul-
une.

Cornmen:s:
Absence of feature 12 implies chai Lhe vessel outlinc is circular to oval (fig. 26) as in Banara
regia (Flacourtiaceae) and the latcwood oU red oaks (e. g,, Que rcusfcdcara - Fagaceae).

Caution: For fossil or archaeological samples, use this tèature only when there obviously is no
distortion from shlinkagc or post-depositional events. Distortion and 'folding' ofthe rays mdi-
cates that the wood has been compressed during burial and that vessel outline probably has heen
aliered.

Fig. 25. Solitary vessel outline angular (feature 12), Stemonurus luzoniensis, x 75. -- Fig. 26.
Outline of vessels rounded (feature 12 absent), Banara regia, x 45.
246 - IAWA Bulietin n.s, Vol. 10(3), 1989
1 AWA List of microscopic features for hardwood identification 247

PERFORATION PLATES

13. Siniple perforation plates

14. Scalarifor'm perforation plates
15. Scalariform perforation plates with 10 bars
16. Scalariform perforation plates with 10-20 bars
17. Scalariform perforation plates with 20-40 bars
18. Scalariform perforation plates with ^ 40 bars
19. Reticulate, forarninate, and/or other types of multiple perforation plates

Definizions:
Simple perforation plate = a perforation plate with a single circular or elliptical opening,
Íg. 27, e. g., Aesculus hippocasranwn (Hippocastanaceae), Entandrophragrna spp. (Meliaceae),
Prerocarpus spp. (Papilionaceae), Zellwva spp. (Ulmaceae).
Scalariform perforation plate = a perforation plate with elongated and parailel openings
separated by one to many mainly unbranched bars, lig. 29, examples follow.
Scalariforrn perforation plates with :^ 10 bars, e. g., Corytus aveilana (Corylaceae),
Goupia spp. (Goupiaccae), Liriodendron tulipifera (Magnoliaccae), Coula edulis (Olacaceae),
Rhizophora mangle (Rhizophoraceae).
Scalariform perforation plates with 10-20 bars, e.g., Betula verrucosa (Betula-
ceae), Altingia excelsa, Liquida,nbar styracflua (Harnamelidaceae), Saco glortis gabonensis (1 lu-
miriaceae), Schi,na wallichii (Theaceae).
Scalariform perforation plates with 20-40 bars, fig.29, e.g.,Cercidiphyllum japo-
nicum (Cercidiphyllaceae), Dicoryphe sripulacea (l-Iamamelidaceae), Nyssa ogeche (Nyssaceae),
Staphylea pinn.ara (Staphyleaceae).
Scalariform perforation plates with ^ 40 bars, e.g., Aexroxicon puncrazum (Aex-
toxicaceae), Hedyosmwn spp. (Chloranth aceae), Dillenia iriquerra (Dileniaceae).
Reticuate perforation plate = a plate with closely spaced openings separated hy wall
portions that are much narrower than the spaces between them, or with a profuse and irregular
branching of wall portions resulting in a netlike appearance, fig. 30, e. g., Didymopanax moro-
toroni (Araliaceae), Iryant hera juruensis (Myristicaceae).
Forarninate perforation plate = a plate with circulas or elliptical openings like a sieve:
the rernaining wall portions can be thicker than ín the reticulate type, fig. 3 1, e. g., Oroxylum
indicum (Bignoniaceae).
Other types = for instance, complex or radiate perforation plates, see comments and figs.
32-35.

Co,nments:
Determine the type(s) of perforation plate fron radial sections or maccrations, preferahly
examine at least 25 vessel elements. For scalariform perforation plates, record ali the fcature
categories thar encompass the range of the number of bars. Feature 14 'scalariforrn perforation
plates' is a general category included to accommodate infortnation from cxisting literature that
indicates whether scalariform perforation platcs are prescnt, but not the number of bars. Feature
14 is to be recorded wi th other appropriate features for bar number (15 –18).
Fig. 27. Simple perforation plates (fcature 13), Aesculus hippocastanum, x 105. - Fig. 28.
Simple perforation plate and scala.riform perforation plate with 2 bars (features 13, 14, 15),
Didynopanax morotoronj, x 115. - Fig. 29. Scalariforrn perforation plate with 20-40 bars
(features 14 and 17), Staphylea pinnata, x 220.— Fig. 30. Reticulate perforation plate (feature
19), Didymopanax ,norototonj x 115. - Fig. 31. Foraminate perforation platc (feature 19),
Oro'.hvn indu ,iun, 115.
248 - IAWA Bulletin n.s., Vol. 10 (3), 1989 IÁWA LisL of inicroscopie features for hardwood idenLification 249

Simple perforations are the most common Lype of perforation plate, and occur in over 80% of
the world's woods (Wheeler et ai. 1986). Most woods have exclusively sirnple perforations,
some have simple perforations together with scalariforin and/or other types of multiple perfora-
tiori plates, and still others have exclusively scaiariform perforalion plates. When more tban one
type of perforation plate is present (fig. 28), ai] types should be recorded and may be uscd to
identify ao unknown (e.g., Didymopanax mororotoni— Araiiaceae, Oxydendron arboreum-
Ericaceae, Fa,us sylvatica - Fagaceae, and Piaranus occidenralis - Platanaceae have both sim-
pie and scalariform plates). In those woods with both simple and scalariforrn perforalion plates,
the narrower vessel elcmcnts and the latewood vessel elements are more likely to have scalari-
form perforatioii piares.
Scalariforrn, reticulate, and foraminate piares forni a continuum, and the larter mo are often
confused in the literature. Reticuiate and forarninate plates are restricted to re]ative]y íew taxon-
oniic groups and are cornbined here. Reticulate perforations frequent]y occur itt combinatiori
with scalariform piares and are an elaboration of that type. !ryanthera (Myristicaceae), Dendro-
panax and Didymopanax (Araliaceae) have scaiariforrn, reticulate, and varied intermediates plus
simpie perforations; Myrceugenia estreilensis (Myrtaceae) has simple and niuldplc plates and the
larter can be variousiy described as irregular-scaiariforni, forarninate, or even reticulate; Mark-
hainia and Oro.xylu.rn (Bignoniaceae) have simple and foraminate platcs.
!ryanthera (Myristicaceae) also lias cornpound scaiariform piares with few coarse bars with
scts of fine secondary bars hetween them, which are often branchcd. Similar examples occur in
Didymopanax marototoni (Araliaceae) and Ternsrroemia serram Theaceae). In these cases, botli
feature 14 (scalariform perforation piares present) and feature 19 appiy. As pointed out by Cari-
•1
quist (1988), the temi 'ephedroid' shouid not be used for foraminate perforations in dicotyle-
lá dons,
Radiate perforation piates (alto feature 19) with a central wall and radiating simp]e and
branched bars extending to the lateral vessel wall are found in Cyiharexylurn myrianthwn (Ver-
benaceae) (Vidal Gomes ex ai, 1989) and Caryocar microcarpum (Caryocaraceae).
v. Other types of multiple perforations may be found in the future and should be recorded as
feature 19, 'reticulate, forarninate, and/or other types of rnultiple perforations'.

34

Fig. 32. Perforation plate obliquely compound scalariforrn with anasiornosing bars (features 14
and 19), Jryanthera paraensis, x 290. Fig. 33. Perforation plate regularly reticulate (feature
19, often occurring together with feature 14, scalariform perforarion p]ates), Iryantherajuru-
ensLs, x 290. - Fig. 34. Perforation plate cornpiex scalariforni and reticu]ale (features 14 and
19), Iryanthera elliptica, x 290, - Fig. 35. Radiate perforation piatc (feature 19), Cyrharexyiwn
myrianrhum, x 300 (SEM).

250 - IAWA Bulietin n. s., Vol. 10 (3), 1989 IAWA List of rnicroscopic features for hardwood identification 251

INTERVESSEL PITS: ARRANGEMENT AND SIZE

20. Intervessel pits scalariform

21. Intervessel pits opposite & .--]=-.
22. Intervessel pits alternate
23. Shape of alternate pits polygonal
Intervessel pit size (alternate and opposite)
24. Minute - ^ 4 im
25. Small — 4-7 im
26. Medium - 7-10 im
27. Large - ^ 10 im
:
28. Range of intervessel pit size (J.Lm)
--; 2

Definitions: Z3(

Intervessel pits = piLs between vessel elements.

Scalariform intervessel pus = elongated or linear intervessel pits arrange.d in a ladder-
like series, fig. 36, e.g., Dillenia reiiculara (Dilleniaceae), Michelia compressa (Magnoliaceae),
Laurelia spp. (Monirniaceae), R/zizophora spp. (Rhizophoraceae).
Opposite intervessel pits = intervessel pits arranged in short to long horizontal rows,
i. c., rows orientated transversely across the Iength of the vessel, figa. 37, 44, C. g- Lirioden-
dron spp. (Magnoliaceae), Nyssa ogeche (Nyssaceae).
Alternate intervessel pits = intervessel pits arranged in diagonal rows, figs. 39, 40, 42,
43, e. g., Aceraceae, Mappia racemosa (Icacinaceae), Leguininosae, Mel iaceae, Salix spp. (Sali-
caceae), Sapindaceae,
Shape of alternate pits polygonal = outline of intei-vessel pits, as seen in surface view
(longitudinal sections), angular and with more than 4 sides, fig.40, e. g., Salicaceae, most Legu-
minosae.
[ntervessel p11 sim (alternate and opposite) = horizontal diameter of a pit chamber at
the broadest point 40

Procedure:
Generaily, surface views of intervessel pits are easiest to find in tangential sections because
radial multiples are the most frequent type of vessel multiple, and so intervessei pits are rnost
frequent in tangenlial walis. When vesseis are in tangential bands and/or clusters, radial sec-
tions also provide surface views of intervessel pits. In woods with (almost) exelusively solitary
vesseis, intervessel pits will be extimely rare, and often not visible in a single longitudinal

Fig. 36 Intervessel pits scalariform (feature 20), Michelia compressa, x 115. - Fig. 37. Inter-
vessel pits opposite (feature 21), Liriodendron tulipifera, x 115. — Fig 38. Intervessel pits sca-
lariform to opposite (features 20 and 21), Ilex laurina, x 350. --- Fig. 39. Intervessel pits
alternate (feature 22), Mappia racernosa, x 112. — Fig. 40. Shape of alternate pits polygonal
(feature 23), Salix sp. Note also features 22 (pits alternate) and 26, 27 (intervessel pits medium
to large); x 290. — Fig. 41. Shape of intervessel pits circular t o oval (feature 23 absent). Note
also plIs opposite (feature 21), Nothofagus moorei, x 300. — Fig. 42. Intervessel pits minute,
less rhan 4 p.m (feature 24), Polyalthia oblongifolia. Note also pits alternate (feature 22), x 300.
— Fig. 43. Vesturesl intervessel pits (feature 29), Terminalia sp., x 825. — Fig. 44. Pits seern- 1
ingly vestured due to presence of soluble deposits (feature 29, vestured pits, ahsent). Note also
pits opposite (feature 21). Ilex cyniosa, x 800.

1
 252 tAWA Builetin ns., Vol. 10 (3), 1989 IAWA List of microscopic features for hardwood identification - 253
section. In such woods, intervessel pit shape and size must be observed in overlapping end wall
portions of vessel elements in a single vessel. However, in woods with vessel multiples, lhe pit
arrangemcnt, shape, and size is best determined frorn lhe middle of lhe larger vessel clements.
Measure ten pits, avoiding exceptionally large or small pits, and record those size classes that
encompass lhe range ofpit size.
Comments:
Alternate intervessel pitting is lhe most common, and opposite and scalariform intervessel
pitting are found in relatively few groups. When alternate pits are crowded lhe outlines of lhe
pits tend to be polygonal in surface view; if alternate pits are not crowded thcn lhe outlines are
usually circular to oval. \Vhen opposite intervessel pits are crowded lhe outlines of lhe borders
tend to be rectangular in surface view. Some spccies have both polygonal and circular to oval
intervessel pit outlines, record featurc 23 present for such woods. Combinations of different
pitting patterns and/or intergrad.ing types occur (e.g., figs. 38, 41, alternate and opposite in
Buxus - Buxaceac, opposite and scalai-iform in Liquidam bar - Hamamelidaceae) and may be
indicated by using combinations of Lhe different pit fcatures.
Pit size can help distinguish between genera within a family and between fainilies, e. g_
many Meliaceae have minute pits, while many members of the Anacardiaceae have large pits.
The most widely used convention for determining pit size is to measure horizontal pit diame ter.
To enable use of existing data this pararneter is included in this feature list. However, within
several taxa, particularly those with some or ali opposite to scalaiiform pits, vertical diameter is a
more constant and diagnostic feature, and it is recommended that this dirnension be recorded in a
description.
Cauzion: Do n'oL mistake vessel-vasiceniric tracheid pitting for intervessei pitting.
VESTURED PITS
29. Vesturcd pits
Definition:
Vestured pits = pits with lhe pit cavity and/or aperture wholl y or partly lined with pro-
jections from lhe secondary ccli wall, fig. 43, e. g., Comhretaceae, Lythraceae, Myrtaceae,
Rubiaceae, most Legurninosae.
Procedure:
Vestures are best viewed in water or glycerin mounts (or SEM). Bleaching is recornmended
50 as to remove encrusting materiais that nlay be mistaken for vestures (fig. 44 shows pseudo-
vestures), i.e., soak sections (or, for SEM observation, wood blocks) in any household bleach
until Lhe section or surface has lost its colour, rinse in water, and finish saniple preparation.
Comments:
Vesturing may occur in intervessel, vessel-ray or vessei-axial parenchyma, intertracheid, or
interfibre pits.
Vestures generaily are characteristic of entire farnilies, or groups within a famiiy. The num-
ber, size, and distribution of vestures varies considerably and these variations may be of diag-
nosuc value (Baiiey 1933; Ohtani eral. 1984; Van Vliet 1978; Van Vliet & Baas 1984). When
intervessel pits are large and Lhe vestures are coarse (e. g., Terminalia spp. - Combretaceae),
vestures are relatively easy to see with an oii-immersion objective of a good compound micro-
scope. But when lhe vestured pits are minute (4 irn or less) as in lhe Apocynaceae or Rubiaceae,
lhe vestures are difficuit to see with a compound microscope, and only clearly visible with a
scanning electron microscope.
VESSEL-RAY PITFING
30. Vessel-ray pits with distinct borders; similar to intervessel pits in size and
shape throughout the ray ceil
31. Vessel-ray pits with much reduced borders to apparently simple:
pits rounded or angular
32. Vessel-ray pits with rnuch reduced borders to apparently simple:
pits horizontal (scalariform, gash . like) to vertical (palisade)
33. Vessel-ray pits of two distinct sizes or types in the sarne ray celI
34. Vessel-ray pits unilaterally compound and coarse (over 10 jm)
35. Vessel-ray pits restricted to marginal rows
Dejlnitions:
Vessel-ray pits = pits between a ray celi and a vessel eicmcnt.
linilaterally compound pita = pits in which one pil abuts two or more smalier pits in the
adjacent celi.
Other features as per descriptors, examples follow.
Vessel-ray pits with distinct borders; similar to intervessel pts in size and
shape throughout the ray cell, figs. 45, 46, e.g., Aceraceae, Leguminosae, Meliaceae,
Jiex aquifoliwn (Aquifoliaceae), Betula spp. (Betulaceae), Camprostemon philippinense (Bom-
bacaceae), Couratari cf. oblongifolia (Lecythidaceae).
Vessel-ray pita with much reduced borders to apparently simple: pita round-
ed or angular, figs. 47, 48, e. g., Elaeocarpus calomala (Elaeocarpaceae), Clinostemon spp.
(Lauraceae), Eucalyptus spp. (Myrtaceae). Popular spp., Salix spp. (Salicaceae).
Vessel-ray pits with much reduced borders to apparently simple: pits hori-
zontal (scalariform, gash-like) to vertical (palisade), figs.49, 50, e.g., Trigonobala-
nus vertici/lata, Quercus spp. (Fagaccae), Atherosperma moschata, Laurelia aromatica (Moni-
miaceae), HorsJieldia subglahosa (Myristicaceae), Syzygiwn spp. (Myrtaceae).
Vessel-ray pita of two distinct sizes or types in the sarne ray cell, figs. 51, 52,
e.g., some species of Erythroxylum (Erythroxyiaceae), Anacolosa spp., Chaunochiton spp.
(Olacaceae), Santa/um spp. (Santalaceae), Planchoneila spp. (Sapotaceae).
Vessel-ray pits unilaterally compound and coarse (over 10 Lm), fig. 53, e.g.,
Michelia champaca (Magnoliaceae), Ceriops spp., Kandelia spp., Rhizophora spp. (Rhizopho-
raceae).
Vessel-ray pits restricted to marginal rows, fig. 47, e. g., Carpinus bezulus (Co-
rylaceae), Aesculus hippocasranum (Hippocastanaceae), Populus spp., Salix spp. (Salicaceae).
Comrnenrs:
Vai-ious combinations of lhe above features may occur and should be recorded. Vessei-ray
pits in lhe body of lhe ray may differ from those in lhe ray margins (e. g., Pala quium gaiato-
xylum - Sapotaceae). Record lhe features for both types of pits.
lia wood has predoniinantly solitary vesseis, comparison of vessel-ray pits with intervessel
pita often is not possibie. li lhe vessei-ray parenchyma pits in such woods are uniform in size
and shape and have borders, then use feature 30; ifnot, any of features 31-35 may apply.
Vessel-axial parenchyma pitting usually resembles vessel-ray parenchyrna pitting, and is
therefore not inciuded as a separate iist of alrnost identical descriptors.
 254 IAWA l3ulletin n.s., Vol. 10 (3), 1989
IAWA List of niicroscopic features for hardwood identification 255

Kffi V!1i

'jpT
Fig. 52. Vesscl—ray pits of two distinct sizos or types iii the sarne ray ce.l1 (feature 33); thc large
pits (arrowcd) resembie perforations, Chaunochiton brevifloruin, x 290. - Fig. 53. Vessel—ray
pits uni]aterally compound and coarse (feature 34), Ceriops raa1 (differentiai interference con-
trast), x .450,

Figs. 45 & 46. Vessel—ray pits with distinct borders, similar to intervessel pits (feature 30). -
45: Courazari cf. obIongfolla, x 290. - 46: Carnpwste,non philippinense, x 75. Figs. 47 &
48. Vessel—ray pits wir.h rnuch reduced borders to apparently simple, pit outline rounded (feature
31), —47: Salix sp. (Salicaccae). Note also feature 35 (vessel—ray pits restrictcd to marginal
rows); x 290. - 48: Elaeocarpus calornala, x 290. - Figs. 49 & 50. Vessel-ray pits with rnuch
reduced borders co apparently simp]e, pits horizontal (gash-Iikc) to vertical (palisade), fc.ature
32. - 49: Pits horizontal, Atherosperrna ,noschata, x 450. - 50: Pics vertical, Trigonobalanus
verticillara, x 290. - Fig. 51. Part of the vessel-ray pits with much reduced borders, and pits of
Iwo distinct sizes or types in the sarne ray ccli (features 32 and 33), I-Iorsfieldia suhglobosa,
x 115.

256 IAWA Builetin n.s., Vol. 10 (3), 1989 IAWA List of rnicroscopic features for hardwood identification 257

HELICAL THICKENINGS

36. Helical thickenings in vessei elements present

37. Hetical thickenings throughout body of vessel element
38. Helical thickenings only in vessel element tails
39. Hetical thickenings only in narrower vessel elements tr

Definitions:
Helical thickenings in vessel etements = ridges on the inner face of the vessel elemeni
wall in a roughly helical pattern. Synonym: spiral thickenings. 1
Other features as per descriptors, exarnples follow.
Helical thickenings throughout body of vessel etement, figs. 54, 55, e.g., Acer
spp. (Aceraceae), Aesculus spp. (Hippocastanaceae), Cyrisus scoparius (Papilionaceae), Prunus
spinosa (Rosaceae), Tilia spp. (Tiliaceae).
Helical thickenings onty in vessel element tails, figs. 56, 57, e. g., Cercidiphyllum
japonicwn (Cercidiphyllaceae), Liquidambar styraczflua (Hamamelidaceae).
Helical thickenings only in narrower vessel elements, e.g., Robinia pseudoacacia f
(Papilionaceae), Ulmus americana (Ulmaeeae). 5--
(54
) h ,4
Commenrs:
Helicai thckenings are rather variable in terms of thickness (fine to coarse), inclination angle
(nearly horizontal to steeply inclined), branching (branched or unbranched), and spacing (dose
to wide). li is recomxnended that observations on these features be included in wood descrip-
tions.
Feature 36 'helical thickenings in vessel elements' is included as a general category to 1) ad-
commodate information frorn existing databases that indicate whether helical thickenings are
present, but not their specific location; and 2) to help with the identification of small wood frag-
ments in which vessel elements have helical thickenings, but because of the small sample size it
cannot be determined whether Lhe helical thickenings are in ali vessel elements, or just in Lhe
narrower ones. Feature 36 should be recorded in combination with other appropriate features
(features 37, 38 or 39) for helical thickenings.
Helical thickenings can also occur in vascular! vasicentric tracheids, and in ground tissue
fibres (feature 64), and very rarely in axial parenchyma.

Caution: Do not confuse coalescent pit apertures with helicai thickenings.

Figs. 54 & 55. Helical thickenings throughout body of vessel elernent (features 36 and 37). -
54: Prunus spinosa, x 290. - 55: cyrisus scoparius, x 220. Figs. 56 & 57. Helical thicken-
ings only in vessel elemeni tails (features 36 and 31), Cercidiphvllurn japonicuin. - 56: x 150. -
57: x 240.

258 IAWA Bulietin n.s., Vol. 10 (3), 1989 IAWA List of microscopic features for hardwood identification 259

TANGENTIAL DIAMETER OF VESSEL LUMINA VESSELS PER SQUARE MILLIMETRE

Mean tangential diameter of vessel lumina 46. !^ 5 vessels per square millimetre
40. ^ 50 jim 47. 5-20 vesseis per square miltimetre
41. 50-100 gm 48. 20-40 vesseis per square millimetre
42. 100-200 im 49. 40-100 vesseis per square millimetre
43. ^- 200 tm 50. ^ 100 vessels per square millimetre
44. Mean, +I . Standard Deviation, Range, n = 51. Mean, +I Standard Deviation, Range, n = x

Procedure:
Vessel diameter is measured in transverse sections. Vesseis are selectcd for measurement
with care not to bias the selection towards the larger or smaller vesseis. The rangential diameter
of the vessei luniina, excluding the wali, is measured at the widest part of Lhe opening. At least
25 vesseis should be measured. Ining-porous woods (feature 3) and woods with 'vcssels of
two distinct diameter classes, wood not ring-porous' (feature 45), only measure and record the
larger size class. Information about tangential diameters of the smailer vesseis wouid be useful
in a description. In semi-ring-porous woods, measure along a radial transect through a growth
ring. For semi-ring-porous woods, it is recommended that more than 25 vessels be measured;
a larger standard dcviation is expected for such woods. Use the category(ies) in which the
mean(s) fali(s).
Comments:
In trees, mean tangential diameters of 100-200 im are more cornrnon than mean tangential
diarneters greater than 200 im or mean tangential diameters less than 50 J.inl. In shrubs, mean
tangential diameters of less than 50 jim are conimon.
Procedure:
Ali vcssels are counted as individuais, e. g., a radial multiple of four would be counted as
four vesseis (Wheeler 1986). Count ali the vesseis in at least tive (and preferably ten) fields of
appropriate size (depending on vessel diameter and distribution), and convert to number per
square rniliimetre, i. e., for woods with small diameter vessels use fields 1 mm x 1 mm or iess;
for woods with large vesseis that are widely spaced use whole fields of view at low magnifica.
tion (e. g., 4 x objective lens). Of the vesseis that are partially in Lhe field of view, only 50% are
included in the count. If vessei frequency is very low, examine enough fields to account for
local variations, and preferably count at least 100 vesseis. Use the categories that include the
total range of vessel frequency.
Cornmenrs:
Vessel frequency is not computed for ring-porous woods, or for woods with their vesseis ia
definite tracts with vascular/vasicentric tracheids, e. g., dendritic pattern as seen in Rhamnu.s
cathartica (Rhamnaceae), or tangential bands as seen in Ulmus (Ulmaceae).

MEAN VESSEL ELEMENT LENGTH

45. Vesseis of two distinet diameter classes, wood not ring-porous
52. ^ 350 l.m
Definirion: 53. 350-800
Vesseis of two distinct diameter classes, wood not ring.porous woods with a 54. ^ 800 im
bimodal distribution of tangential diameters of vessel lurnina, fig. 58, e. g., Actinidia spp. (Acli- 55. Mean, +1— Standard i)eviation, Range, n = x
nidiaceae), Capparis maroniensis (Capparidaceae), Derris hylobia Papilionaceae), Serjania
subdenrata (Sapindaceae), Congea ro,nenrosa (Verbenaceae). Procedure:
Measure the whole iength of each vessel element from one tail end to thc other, preferably in
Comrnen ts: a maceration. AL least 25 vessel element lengths are measured to derive the mean and range. Use
Vines and xerophytes often have vesseis of two distinet diameter classes (Carlquist 1985; the categoly(ies) in which the mean(s) fali(s).
Baas & Schweingruber 1987).

TYLOSES ANO DEPOSITS IN VESSELS

56. Tyloses common

57. Tyloses sclerotic

Definitions:
Tyloses common = outgrowths from an adjacent ray or axial parenchyma ccli through a pit
in a vessel wall, partially or completely blocking the vessel lumen, and ofcQmmon OccUlTence
260 - - IAWA Builetin ns, Vol, 10 (3), 1989 IAWA List of rnicroscopic features for hardwood identification -- - -- 261

(except in outer sapwood), figs. 59, 60, e. g_, Anacardium occidentalis (Anacardiaceae), Acan-
rhopanax spp. (Araliaceae), Cercidiphyllum japonicum (CercidiphyUaceae), Eucalyptus acme-
niaides (Myrtaceae), Sti-ombosia pustulata (Oiacaceae), Robinia pseudoacacia (Papiionaceae).
Selerotie tyloses = tyloses with very thick, multilayered, lignified walis, figs. 61, 62,
e, g., Chaetocarpus schoni burgki anus, Micrandra spruceana (Euphorbiaceae), Cantleya cornicu-
lata (Icacinaceae), Eusideroxylon zwageri (Lauraceae). Brosimum guianense (Moraceae).

Procedure:
In ring-porous woods, it is best to examine the earlywood vesseis for tyloses because ty10es
are often absent froni small diarneter latewood vesseis. Avoid sapwood when determining the
presence of tyloses or gums.

Commenrs:
Tyloses may be few or many, ranging from ali vesseis fihled with many tyloses to a few ves-
seis with a few tyloses. Feature 56 applies only when tyloses are not of sporadic occurrence.
Tyloses may be thin-walled or thick-walled, pitted or unpitte, with or without starch, crystals,
Ç '..-sVl resins, gums, etc. Such information should be recorded in a description.
Woods with selerotie tyloses usually have thin-walled tyloses as well, and both features 56
and 57 tnay apply Some woods may have both tyloses and guni deposits (feature 58), and both
1 features 56 and 58 may apply.

Caution: Absence af tylases is not diagnostic Do um code lraumatic tyloses such as occur in
wound heartwood and be careful rim to confuse tyloses with foamy deposits, masses of fungi,
ar other deposits.

fl L' 58. Gums and other deposits in henrtwood vesseis

(//l Cornments:
In cross sections, deposits appear to completely fihi some vessel lumina (fig. 63); in longitu-
dinal sections, deposits often appear to collect at the end of vessei elements. Deposits often can
be seen more clearly by exarnining the woods with a hand lens; sectioning and rnounting tech-
fiques may remove some of the deposits.
-;; 'Gums and other deposits' includes a wide variety of chemical compounds, which are vari-
ously coloured (white, ye]iow, red, brown, black). Ia a description it is appropriate to indicate
- their abundance and colour. See Hillis (1987) for more informadon on the chemistry of depo-
-
,L sits.

Cautíon: Use feature 58 positively oniy. Do not confuse masses of whitish fungi, which ma y be
packed in a vessel cavity, ar sclerotic tyloses with deposits.
- 1-.
.1

H.
Fig. 58. Vesseis of two disdnct diameter classes, wood not ring-porous (feature 45), Serjania
subdentata, x 35, - Figs, 59 & 60. Tyloses common (feature 56). – 59: In transverse section,
Anacardiurn occidentalts, x 220. - 60: Ia tangential secdon, Robinia pseudoacacia, x 140. —
-: (1 Figs. 61 & 62. Tyloses sclerotic (features 56 and 57), Cantleya corniculata. – 61: Transverse
62 scction, x 75. – 62: Tangential section, x 75. — Fig. 63. Gums/dcposits in heartwood vesseis
(feature 58), Physocalymma scaberrimum, x 110.

E-

262 IAWA Bulietin n.s., Vol. 10 (3), 1989 IAWA List of microscopie features for hardwood identification 263

WOOD VESSELLESS

59. Wood vesselless

Definition:
Wood vesselless = wood without vessel elements, cornposed only of imperforate trache-
ary elements and parenchyma, figs. 64, 65, e. g., Amborellaceae, Tetracentraceae, Trochoden-
draceae, Winteraceae.

Comnients:
Vesselless dicotyledonous woods are relatively uncornmon and are distinguished from conif-
erous wood by tall multiseriate rays. For vesselless woods, it is not necessary to cocle for type
of imperforate tracheary elements (fibres or vascular/vasicentric tracheids) and impossible to
code for vessel features.

TRACHEIDS AND FIBEES

60. Vascular/vasicentric tracheids present

Deflnitions, .1: uI
Vascular tracheids = imperforate edis resembling in size, shape, pitting, and wall orna-
mentation narrow vessel elements and intergrading with the lattcr, figs. 66, 67, e.g., Sambuctis
4 t
nigra (Caprifoliaceae), Sop hora arizonica (Papiionaceae), Phellodendron amuren.se (Rutaceae),
Lycium europaewn (So!anaceae), Celtis occidenralis (Ulinaceae).
Vasicentric tracheids = imperforate celis with numerous distinctly bordered pits in their
radial and tangenhial walis, preserir around the vesseis, and different from ground tissue fibres,
often, but not always of irregular shape, fig. 68, e. g., Castanea spp., Quercus spp. (Fagaceae),
rnany Shorea (Dipterocarpaceae) and Eucalyprus (Myrtaceae) species.

Comments:
s1
Vascular tracheids often occur in association with extensive vessel rnultiples or clusters,
especially in the latewood. A very thorough search of macerations will reveal their presence iri
tnany species. 1-lowever, for wood identification purposes use rhein only when they are com-
monly present. The intergradation of vascular tracheids with narrow vessel elements implies thar 67
there are some celis with a single, often very small, perforation. Some anatomists would prefer
to cail such edis vascular tracheids, others would prefer to cail them narrow vessel elements, Figs. 64 & 65. Wood vesselless (feature 59), Trochodendron aralioides. - 64: Transverse
probably terminating a ves.sel. Because they have a perforation such cells are best referred to as section, x 38. - 65: Tangential section, x 38. - Figs. 66-68. Vascular/vasicentnc tracheids
vessei elements; tracheids are imperforare edis. (feature 60). 66: Radial section, vessei elements surrounded by vascular tracheids (note also
The IAWA Glossary (1964) includes shortness and irregular form in the definition of vasi- helicai thickenings in vessel elements, features 36 and 37), Phellodendron a.'nurense, x 290, -
centric tracheids, but these criteria do nor always apply (e.g., iii Eucalyptus spp., cf. lhe 1987). 67: Maceration, vessel element flanked by two vascular tracheids, Celtis occidentalis, x 300. -
Since vascular tracheids are often intermixed with vessels (i.e., in a vasicentric position) in 68: Radial section, vasicentric tracheids, Castanea sariva, x 170.
many taxa, they can also be considered as vasicentric iracheids.

264 IAWA Builetin n.s., Vol. 10 (3), 1989 IAWAList of microscopic features for hardwood identification 265

GROUND TISSUE FIBRES

61. Fibres with simple to minutely bordered pits

62. Fibres with distinctly bordered pits
63. Fibre pits common in both radial and tangential walis

Defini rions:
Fibres with simple to minutely bordered pits = fibres (libriform fibres) with simple ii
pits or bordered pits with the chambers less than 3 tm in diameter, flgs. 71, 72, e. g., Swietenia
spp. (Meliaceac), Inga spp. (Mimosaceae), Fraxinus spp. (Oleaceae), Populus spp. (Salicaceae).
•
Fibres with distinctly bordered pits = fibres (or fibre-tracheids or ground tissue isa-
cheids) with bordered pits with chambers over 3 p.m in diameter, figs. 69, 70, e.g., 1/ex spp. • 1.
(Aquifoliaceae), Dilienia spp. (Dileniaceae), Illiciu.m spp. (llliciaceae), Xanrhophyllum spp.
(Polygalaceae), Carneilia spp. (Theaceae).
Fibre pits common in both radial and tangential walis = fibre pits, either bordered
or simple, common in radial and tangential walis, e. g., [lex spp. (Aquifoliaceae), Dilienia spp.
(Dilleniaceae), 11/id um spp. (Illiciaceae), Xanthophyllum spp. (Polygalaceae), Clematis vitalba U4
(Ranunculaceac), Carne/lia spp. (Theaceae).

Procedure: llt1 -
Determine the nature and distribution of fibre pits only in longitudinal (radial and tangential)
sections, because in cross section many fibre walis are not strictly radial or tangendal. Both lon- '4..'.
gitudinal and cross sections are suitable to determine if the pits are bordered or (almost) simple.

Commen:s:
The feature 'fibres with distinctly bordered pits' partly overlaps with the descriptors 'tracheids'
1
sensu Bailey (1936) and Carlquist (1986a, 1986b, 1988) and 'fibre-tracheids' sensu Baas
(1986). It usually coincides with 'fibre pits common in both radial and tangential walis' (feature Ls'
63).
The following combinations are of very sporadic occurrence: 1) fibre pits not distinctly bor- Li
dered, ie., pit chambers less than 3 j.un or pits simple, feature 61 present, and pits comrnon in
radial and tangential walls, feature 63 present, e. g, Capparis spinosa (Capparidaceae), Nyctan-
thes arbor-tristis (Oleaceae), Vitis vin(fera (Vitaceae); 2) fibres with distinctly bordered pits, fea-
ture 62 present, and pits mainly restricted to the radial walis, feature 63 absent, e. g.. Elaeagnu.s
angu.safolia (Elaeagnaceae).
Two types of fibres with respect to wall pitting (both features 61 and 62 present) may occur
(e. g, some species of Vaccinium - Ericaceae).
I1
Fibres with simple to minutely bordered pits (feature 61 present), mainly confined to the
radial walis (feature 63 absent) are libriform fibres in the deflnition of Baas (1986) or libriform
Figs. 69 & 70. Fibres with distinctly bordered pits (feature 62), common in both radial and t:m-
fibras and/or fibre-tracheids in the deflniuon ofCarlquist (1986a, 1986b, 1988).
gential walls (feature 63), tangential sections. - 69: lilicium carnbodianu,n, x 230. - 70: Xant/i
The terms libriform fibras, fibre-tracheids, and 'true tracheids' have been deliberately avoid-
phyllurn lanceatu,n, x 230. - Fig. 71. Fibrcs with simple to minutely bordered pOs (feature
cd as descriptors in this list because there is no consensus on their deflnitions.
61), tangential section, phase-contrast, Populus sp. (Salicaceae). Note also .scarcity ofpits in
tangential walis (feature 63 ahsent); x 410. Fig. 72. Fibrcs with simple to minutely bordered
pOs (feature 61), common in both radial and tangcntial walis (feature 63). Tangential seclion,
phase-contrast, Clernatis viralba, x 410. - Figs. 73 & 74. Helical thickcnings iii ground tisue
fibres (fet:re 64). - 73: 1/ex cjnerea. x 850. 74: JIe.ï chinensis, x 850.
Er-

266 IAWA Balietin n.s., Vol, 10 (3), 1989 1 AWA List of rnicroscopic features for hardwood identification 267

64. Helical thickenings in ground tissue fibres

Definition.-
Helical thickenings in ground tissue fibres = as per feature descriptor, see definition
of helical thickenings, feature 36, tigs. 73, 74, e.g., Euonynus europaeus (Celasti-accae), Ha-
name1isjaponica (1 larnamelidaceae), Cercocarpus ledifolius (Rosaceae), and tcmperate species
of 1/ex (Aquifoliaceae) and Syringa (Oleaceae).

Com ments.
Fibres with helical thickenings usualty occur in woods that also have he]ical thickenings in
the vessel elements. However, the opposite is not true, Le., many species with helical thickcn-
ings in their vessel elernents do not have helical thickeriings in the ground tissue fibres. Helical
thickenings are rnuch more cornmon iri fibres with distinctly bordered pits than in tibres with
sinnple to rninutcly bordered pits. They also occur more frequently in tempera te woods thari in
tropical woods,

SEPTATE FIBRES AND PARENCHYMA-LIKE FIBRE BANDS

65. Septate fibres present

66. Noriseptate libres present
67. Parenchyma-Iike fibre bands alternating with ordinary tibres

Definitions:

Septate fibres = fibres with thin, unpitted, transverse wall(s), flgs. 75-78, e.g., Spondias
mombin (Anacardiaceae), A ucournea kíaineana (Burseraceae), Buchenaia capitara (Combreta-
ccae), E!aeocarpus spp. (Elacocarpaceae), Aglaia spp. (Meiaeeae), Virex orinocensis (Verbena-
ceae).

Nonseptate fibres = tibres without septa.

Parenchyma-like libre bands alternating with ordinary fibres = tangential bands
of relatively thin-walled fibres alternating with bands oí thicker-wallcd fibres, fig. 79, e. g.,
Cassine spp., Mayxenus obtusifoila (Celastrace ae), Dubauria spp. (Compositac), Lagersiroemia
tomernosa, Physocalymma spp. (Lythraceae), Cupania americana (Sapindaceae).

Com menu:
Septa are formed after lhe secondary fibre waJls have been deposited; they therefore do not
extend to Lhe compound middle larnellae between adjacent fibres, Septa are usua]Iy unlignified
and veiy thin (cf. Pararneswaran & Liese 1969).
In some woods, ali fibres are septate (feature 65 present, feature 66 absent), e.g., fig. 76,
Lainea welwirschii, Spondias niombin (Anacardiaceae), Canariwn schwemnfiirthii (Bursera-
ceae). In other woods, both septate and nonseptate flbres occur together (features 65 and 66
Figs 75-77. Septate fibres presem (feature 65). —75 & 76: Ali iibres septane, severa] septa per
boih present), eg., fig. 78, Buchenavia capirata (Conibretaceae), E/aro carpus spp. (Elaeocarpa- fibre, Aucouniea klaineana, - 75: x 290 —76: x 75 - 77: Many septa per fibre, Agia/a lirtora/is.
ceae), Swie:enia macrophylla (Meliaceae). The septate fibres rnay then either be scaitered irregu- x 150 Fig, 78. Septate (arrows) and nonseptate fibres presem iii lhe sarne sanipie (features
Iarly, situated near the vesseis or the rays (feature 67 absent), or arranged in tangential bands 65 and 66), Swietenja inacrophyIla, > 115. - Fig. 79L Parenchvtna-iike (ihi'e bands (feature
(feature 67 presem). 67). Phys cah'nnma .scabcrrimu,n, x 45.

268 IAWA Butietin n.s., Vol. 10 (3), 1989 IAWA List of microscopic features for hardwood identification 269

The fibres of parenchyma-like fibre bands (featore 67) are usuaily septate; lhe ordinary fibres
they alternate with may be nonseptate (feature 66) as in Cassine maurocenia, Maytenus obru-
sifolia (Celastraceae), or septate as in Lagersrroernia tornentosa, Physocalymma scaberrimurn
(Lythraceae). In lhe laner case, features 65 and 67 are presem, feature 66 is ahsent. •
, .r.
Fibre septation is coded independently of fibre wall pining (features 61-63).
Thc number of septa per fibre can vary from 1 to many. This number can be taxon specific :.
(e.g., Van Vliet 1976b), and so lhe average number of septa per taxon shouid be given in a 1
description.

Caurions: Do not confuse tom ccli wali fragments, ceil wail deformations, gum deposits, fungal 4• -_'_ 1-.. 1
4.
_,1i_ •' .,.
.
hyphac, or tyloses in fibres (seen in some Magnoiiaceae, Lauraceae) with septa. Sometimes
parenchyma strands can also easily be confused with septate fibres.
•-
Avoid tension wood, because gelatinous fibres are nonseptate.
'• .1
(
r a
,
FIBRE WALL THICKNESS 1 •
•1
68. Fibres very thin-walled 8'
69. Fibres thin- to thick-walled ' •''4

70. Fibres very thick-walled

DeJinitions:
Fibres very thin-wailed = fibre lumina 3 or more times wider than lhe double wall thick-
ness, fig. 80, e. g., in Bursera simaruba (Burseraceae), Te:ra,neles nudzflora (Datiscaceae),
Neubergia corynocarpa (Loganiaceae), Ti/ia japonica (Tiliaceae).
Fibres Ihin- to thick-walled = fibre lumina less than 3 limes the double wall thickness,
and distinctly open, fig. 81, e, g., in 1/ex spp. (Aquifoliaceae), Michelia compressa (Magnolia-
ceae), Salix alba (Salicaceae).
Fibres very thick-walled = fibre lumina a]most compietcly closed, fig. 82, e.g., in
Goupia glabra (Ceiastraceae), Lophira spp. (Ochnaceae), Strombosia pusxulwa (Olacaccac).
Kr/4giodendronferreurn (Rhainnaceae), Rhizophora tnangle (Rhi zophoraceae).
Commen:s:
Measurement of lhe actual thickness of fibre walls usually invoives an amount of work ou[ of
ali proportion to lhe iimited diagnostic value of the figure obtained. Therefore, lhe classes for
fibre wall thickness are based on lhe ratio of lumen to walt thickness (Chattaway 1932). The
ratio proposed is that of lhe width of lhe lumen to lhe combined thickness of lhe walis between it
and lhe lumen of lhe next ccli as viewed in cross section. When celis are flattened radiallv the
lumen becomes oval and will give a different ratio with lhe wail according to whether it is meas-
ured radiaily or tangentially; lhe radial measurement is suggested.
Chattaway (1932) proposed four categories; three are used here. Feature 68 roughiy corre-
sponds to her category very thin'-walied; feature 69 includes her two categories 'thin' and
'thick'; feature 70 is identical to her category 'very thick'.
Fibre wall thickness in many species is variable and there may be more than one category of
fibre wall thickness in a species.
Fig. 80. Fibre.s very thin-walled (feature 68), Neuburgia corynocarpa, x 290. - Fig. 81. Fibres
thin- to thick-walled (feature 69), Michelia compressa, x 290. - Fig. 82. Fibres very thick-
walled (feature 70), Rhizophora rnangle, x 290.
Cauzions: In woods with distinct growth rings, fibre wail thickness changes throughout lhe
growth ring, and rnay be particularly thick aI lhe end of lhe growth ring. When describing fibre
wali thickness, do ,tot consider these iast latewood fibres.
Also, do nol dcscribe gelatinous fibres (—tension wood fibres), which usual]y liave thick
walls with an unlignified gciatinous iayer,
MEAN FIE3RE LENGTI IS
71. 900 j.un
72. 900-1600 j.irn
73. ^ 1600 j.tm
74. Mean, +I Standard Deviation, Range, n = x
Procedure:
Use macerations of mature trunk wood, and measure lhe length of at lcast 25 fibrcs to de
termine lhe mean, range, and standard deviation. Use lhe category(ies) in which lhe mear-i(s)
fali(s). For woods with distinct growth rings, sample from lhe middie of lhe growth ring. Be-
cause of lhe importance of ccli length in wood quality studies, a variety of rnethods have bccn
developed to insure random selection of edis for measurement. It is recommended that one of
these rncthods be used (Dodd 1986; Hart & Swindel 1967). There are very few woods in which
fibre len h :In Tnca1Ired acc uratclv from sccion s. so such a method n O rLc( ,rnrncndcd.
 270 271
- IAWA Builetin n.s., Vol. 10 (3), 1989 lAWAListof microscopic features for hardwood idenrification

AXIAL PABENCHYMA

General commenls:
When identifying an unknown, use the most obvious type of parenchyma pattern first and
then the less evident type or types. Be sure to use a broad ficld of view when deterrnining the
predorninant parenchyma pattern(s) from the n-ansverse section. Various combinations of the
three general types (Apotracheal, Paratracheal, and Banded) described below may be prescrit in a
given wood.

75. Axial parenchyma absent or extremely rare

Definizion:

Axial parenchyrna absent or extremely rare = as per feature descriptc,r, fig. 83, e.g.,
Berberidaceae, Punicaceae, Violaceae, Homaliu,nfoetidwn, Sconellia coriacea (Hacourtiaceac),
Sonneratia spp. (Sonneratiaceae).

Comment:
li is necessary to study longitudinal sections in combinajion with transverse sections to be sure
axial parenchyma is absent or extremely rare (i.e., very difficult to find; only a few stands per
section). This feature may be used in combination with 'axial parenchyma scanty paratracheal'
(feature 78) and/or 'axial parenchyma diffuse' (feature 76), if, despite the scarcity of parenchy-
ma strands, the distribution is clear.

APOTRACHEAL AXIAL PARENCFIYMA

76. Axial parenchyma diffuse

77. Axial parenchyma diffusein.aggregates

Dejinirions:

Apotracheal axial parenchyma = axial parenchyma not associated wth the vessels.
Axial parenchyma diffuse = single parenchyma sirands or pairs of strands distributed
inegularly among the fibrous elements of the wood, fig. 84, e. g., Aspídosperrna polyneuron
(Apocynaceae), A/nus glutinosa (Betulaceae), Goupia glabra (Celastraceae), Cornus mas (Cor-
naceae), Apodytes di,nidiata (Icacinaceae), Crataegus spp. (Rosaceae), Santalum album (San-
talaceae).

Axial parenchyma diffuse-in-aggregats = parenchyma strands grouped into short

discontinuous tangential or oblique lines, fig. 85, e.g., Durio spp. (Bombacaceae), Hura crepi-
tans (Euphorhiaceae), Ongokea gore, Strombosia pusrulata (Olacaceae), Agonandra brasiliensis
(Opiliaceae), Dalbergia sreven.sonii (Pap ilionaceae), Prerospermum spp. (Sterculiaccae), Tilia
spp. (Tiiiaceae).
Fig. 83. Axial parenchvnia absent or extremely rare (featurc 75), Homaliu,nfoezidum, x 75. -
Fig. 84. Axial parenchyma diffuse (feature 76), Alnus glutinosa,x 115. - Fig. 85. Axial
parenchyma diffuse-in-aggregates (feature 77), Agonandra brasiliensis, x 45. - Fig. 86. Axial
parenchyma scanty paratracheal (feature 78), and aiso scanty diffuse (feature 76), Dilienia pul-
cherrima,x 115.
 272 IAWA Builetin ris., Vol. 10(3), 1989 IAWALisL of microscopic features for hardwood identification 273

Co,nmenrs:
Because there is a continuous range from parenchyma extrernely rare, diffuse, diffuse-in-
aggregatcs, to parenchyma in narrow bands (feature 86) or scalariform (feature 88), for some
taxa it wil] be necessary to record more than one feature for apotracheal parenchyma. Diffuse
and diffusc-in-aggregates frequently occurincombination. Record (1944) referred to diffusc-in-
aggregates parenchyrna as reticulate. This list does not follow that usage, but uses rcticulate to
describe a type of barided parenchyma, sce feature 87.
Caurions: Although by defiuiition apotracheal parcnchyma is not associated with vesseis, woods
with abundant diffuse or diffuse-in-aggregate parenchyma rnay exhibit several strands touching
ihe vessels. Such random contacts should not be recorded as parairacheal parenchyma.
Apotracheal diffuse parenchyma sornetimes occurs primarily near the rays ('ray adjacent
parenchyrria' of Carlquist 1988), and shou]d not be confused with sheath celis in rays (feature
110),

PARATRACHEAL AXIAL PARENCHYMA

78. Axial parenchynia scanty paratracheal

79. Axial parenchyma vasicentrie
80. Axial parenchyma aliform
81. Axial parenchyma lozeuge-aliform
82. Axial parenchyma winged-aliform flii .tiIJ(
83. Axial parenchyma confluent
84. Axial pa renchyma unilateral pa rat racheal

Definitions:
Axial parenchyma paratracheal = axial parenchynia associated wih the vesseis or vas-
cular tracheids; Lypes of paratache.al parenchyma are scanty paratracheal, vasiccntric, aliforrn
(subtypes: lozenge-aliform, winged-a]ifonn), confluent, and unilateral paratracheal.
Axial parenchyma scanty paratraclieal = occasiojial parenchyma celis associated with
the vesseis or an incornplete sheath ofparenchyma around Lhe vesseis, lig. 86, eg. Pistacia vera
(Anacardiaceae), Scierolohium spp. (Caesalpiniaceae), Dillenia puicherrima (Dillerriaccae), Eiy-
throxylu,n ,nannii (Erythroxylaceae), Laurus nobilis (Lauraceae).
Axial parenchyma vasicentric = parenchyma celis forining a complete circular to oval
sheath around a solitary vessel or vessel multiple, Ílgs. 87, 88, e. g. Tachigali myr,necophylla
(Caesalpiniaceae), Octomeles sunwrana (Datiscaceae), Phoebe porosa (Lauraceae), Khaya gran-
d(foliola (Mcliaceae), Anadenanrhera spp., Ent-erolobiwn cyclocarpwn, Pipradeniastru,n africa-
num (Mimosaceae), Olea europaea (Oleaceae).
Axial parenchyma aliform = parenchyma surrounding or to one side of the vessel and
with lateral extensions. For examples see Lhe two subtypes below.
Figs. 87 & 88. Axial parernhyrna vasjcentric (featurc 79). - 87: Parenchvuia sheath broad,
Axial parenchytna lozenge-aliform parenchyma surrounding or to orie side of the Pip fade niastrum africanum, x 26. - 88: Parenchyma sheaih narrow, Khavn grandifoliola, x 45.
vessels with lateral extensions forming a diamond-shaped outline, fig. 89, e.g., Albizia lebbek. - Fig. 89. Axial parenchyma lozenge-aliforrn (features 80 and 81). Microberlinia brazzavillen-
Parkia gigantocarpa (Mimosaceae), Arlocarpus chaplasha (Moraceae), Microberlinici brazzavil- sts, x 29. - Fig. 90. Axial parenchyma winged-aliform (features 80 and 82), Brosimu.m rubes-
lensis, Ormosia fiava, Vataireci spp. (Papilionaceae), Quaiea rosca (Vochysiaceae) cens, x 45,
 274 - IAWA Bul]etin n.s., Vol. 10 (3), 1989 LAWA List of nhicroscopic_features for hardwood identification 275

Axial parenchyma winged-aliform = parenchyma surrounding or to one side of the

vesseis with the ]ateral extensions being elongated and narrow, fig. 90, e, g. Jacaranda copaia
(Bignoniaceae), Termina/ia superba (Combretaceae), Brosimum spp. (Moraceae), Quassia ama-
ra (Simaroubaceae), Gonystylus spp. (Thymelaeaccae).
Axial parenchyma confluent = coalescing vasicentrie or aliform parenchyrna surround-
ing or to one side of two or more vesseis, and often forming irregular bands, figs. 91, 92, e. g.,
Kigelia africana (Bignoniaceae), Caesalpiniaferrea, Peito gyne confertifiora (Caesalpiniaceae).
Marmaroxyion racemosum, Parkia pendula (Mimosaceae), Ch/orophora tinctoria (Moraceae),
Bowdichia nitida,Vatairea guianensis (Papilionaceae).
Axial parenchyrna unilateral paratracheal = paratracheal parenchyma forming semi-
circular hoods or caps only on one side of the vesseis and which can extend tangentially or
obliqucly in an aliform or confluent or banded pattern, fig. 93, e.g., Aspidosperrna desman.
thum (Apocynaceae), Caraipa grandiflora (Borinetiaceae), Peito gyne conferíiflora (Caesalpi fia-
ceae), Mammea bongo (Guttiferae), Dilobeia thouarsii (Proteaceae).

Comments:
Scanty paratracheal includes what has been described iti the literature as iricomplete vasicen-
inc.
Some woods have vasiccntric, aliform, and confiuent paratracheal parenchyma. ConflueLu
often intergrades with banded and may be recorded or used in combination with features for
hand width (85-86). Feature 80, 'Axial parenchyma alifornf, is included as a general catcgory
to 1) describe those woods that clearly have aliforni parenchyma, but in which it is difficult to
decide whether it is lozenge- or winged-alifonn, and 2) allow use ofinformation from the litera-
ture that does not differcntiate between these two types. Features 81 and 82 are used in combi-
nation with frature 80.
Feature 84 'unilateral paratracheal parenchyma' is used in cornbination with aliform and/or
confluent when the unilateral parenchynia extends lateraily or obliquely. Unilateral includes both
abaxial and adaxial because generally ii is not possible to discinguish between the two in a wood
fragment.
Woods with several types of paratracheal parenchyma co-occurring and/or intergrading have
been assigned the general descriptor 'parenchyma predominantly paratracheal' by several
authors (fig. 94).

Caution: Vasicenti-ic/vascular tracheids are often thinner-wal]ed than ground tíssue tlbres, and
in cross sections may be confused with axial parenchyma. Examine longitudinal sections to
detennine whether vasicentric/vasctdar tracheids or axial parenchyma surrounds the vesseis,

1-'igs. 91 & 92. Axial parcnchyrna conflucnt (frature 83). —91: Parkiapen(Iuia arrowhead, note
also features 80 and 81, parenchyma ]ozenge-aliforrn), x 36, —92: Peltogyne confertiflora (note
also feature 84, unilateral parenchyma), x 45. Fig. 93. Axial parenchyma unilateral para-
tracheal (feature 84), Caraipa grandiflora, x 45. - Fig, 94. Axial parcnchyma prcdominantly
paratracheal, Garcinia iwissirna (note also features 80, alifonii, and 83, confluent), x 45.
 276 - IAWA Bulietin n.s., Vol. 10 (3). 1989
IAWA List of microscopic features for hardwood identification 277

BANDED PARENCHYMA

85. Axial parenchyma bands more than three cells wide o

86. Axial parenchyma in narrow bands or tines up to three cells wide
87. Axial parenchyma reticulate
88. Axial parenchyma scalariform
89. Axial parenchyma in marginal or in seemingly marginal bands

Definitions:
Parenchyma bands more than three edis wide = as per feature descriptor, fig. 95,
e.g., Dicorynia paraen.sis (Caesalpiniaceae), Entandrophragma candoliei (Meliaceae), Ficus re-
tusa (Moraceae), Lophira alata (Ochnaceae), !Jasyloxylon brasiliensis (Sterculiaceae), Erisma
uncinarwn (Vochysiaceae).
Parenchyma in narrow bands or lines up to three cells wide = as per feature de-
scriptor, figs. 96, 98, e.g., Dialium guianensi.s (Caesalpiniaceae), Endospermum malaccensis
(Euphorbiaceae), Bertholletia excelsa (Lecythidaceae), Dysoxylum frasera num (Meliaccae),
Aurranelia congolensis (Sapotaceae), Hanwa klaineana (Simaroubaceae).
Parenchyma reticulate = parenchyma in continuous tangential lines of approxirnately the
sarne width as the rays, regularly spaced and forming a network with them. The distance be-
tween the rays is approxirnately equal to the distance between the parenchyma bands, lig. 97,
e. g., Cleistopholjs spp. (Annonaceae), Diospyros discolor (Ebenaceae), Bertholletia excelsa,
cariniana spp., Couratari guianensis, Eschwei lera spp. (Lecythidaceae).
Fig. 95. Axial parenchyrna iii hands more than three edis wide (feature 85), Ficus retusa, x 45.
Parenchyma scalariform = parenchyma in fairly regularly spaced fine lines or bands, ar- - Fig. 96. Axial parenchyma in narrow bands or lines up to three edis wide (feature 86).
ranged horizontally or in arcs, appreciably narrower than thc rays and with them producing a .4utranella congolensis, x 45.
ladder-Iike appearance in cross section. The distance between the rays is greater than the distance
between parenchyma bands, figs.99, 100, e.g.,Anisophyllea spp. (Anisophylleaceae), Onycho.
petalum sp. and most other Annonaceae, Cardwellfa sublimis, Embothrium mucronatum (Pro-
teaceae), Rhopalocarpus spp. (Rhopalocarpaceae).
Parenchyma in marginal or in seemingl y marginal bands = parenchyma bands
which fonn a more or lcss continuous Iayer of variable width at the margins of a growth ring
or are irregularly zonate, figs. 101, 102, c. g- Jntsia hijuga (Caesalpiniaccae), Juglans regia
(Juglandaceae), Cryptocarya moscJzaza (Lauraceae), Liriodendron zulipifera, Michelia compressa
(Magnoliaceae), Cedrela spp., Swierenia spp. (Meliaceae), liorsfieldia subglobosa (Myristica-
ceae).
Comments:
Parenchyma bands may be mainly independent of the vesseis (apotracheai), definitely asso-
ciated with the vesseis (paratracheal), or both. Bands may be wavy, diagonal, straight, contiflu-
ous or discontinuous (the latter often intergrading with confluent). The nurnber of bands per mm
varies and may be useful as a diagnostic feature in some groups. Bands over three celis wide are
visible to the unajded eye. For woods with reficulate, scalariform, or marginal parenchyma, the
band width (either fé-atures 85 or 86) also should be recorded.
In the past, some anatomists (e. g., Record 1944) have used the term reticulate for abundant
diffuse-in-aggregates parenchyma with numerous short interrupted lines.
Sometimes marginal parenchyma bands are associated with axial interceliular canais. In some
temperate woods there are discontinuous bands/lines of parenchyma at the growth ring margin;
this condition also should be described as 'marginal'. Marginal parenchyma includes terminal
and initial parenchyma, and seemingly marginal includes what has been called irregular zonate
bands.
 Iiw U:i

278 IAWA Builetin n. s., Vol. 10(3), 1989 IAWA Listof microscopic_features for hardwood identification - - 279

£
z k1am !Wg -1

ÉO,
b4W
5-.
ti AI Ii k

Figs. 101 & 102. Axial parenchyma in marginal or seetningly marginal hands (feature 89). -
101: Inisia bijuga, Note also features 80, 81 (lozenge-aliform parenchyma), 83 (confluent), and
86 (narrow bands); x 29. - 102: Michelia compressa, x 45.
ij

, t?f td2

A.

Fig. 97. Axial parenchyrna reticulate (feature 87), í3ertholleria excelsa (note also feature 86,
parenchyma in narrow bands), x 45. - Fig. 98. Axial parenchyrna intermediate between
reticulate and scalanform (featurcs 87 and 88 variable), Couratari guianensis, x 45. - Figs. 99
& 100. Axial parenchyrna scalariforrn (feature 88). - 99: Onychopetalum sp., x 45. - 100:
Parenchyma bands also 'festooned', Cardwellia sublirnis, x 15.
 281
280 IAWA Bulietin n.s., Vol. 10 (3), 1989 IAWA List of microscopicfeatures_for hardwood identification

AXIAL PARENCHYMA CELL TYPE/STRAND LENGTIT

90. Fusiform parenchyma cells

91. Two edis per parenchyma strand
92. Four (3-4) celis per parenchyma strand
93. Eight (5-8) edis per parenchyma strarid
94. Over eight ceiis per parenchyma strand

Definition.ç:

Fusifoi-m parenchyma = parenchyma celis derived from fusiform cambial initiais without
subdivisions or tip growtli. In shape they resembie a short fibre, fig. 103, e.g., Capparis spp.
(Capparidaceae), Aeschynornene elaphroxvlon, Erythrina spp., Lonchocarpus spp. (Papiliona-
ceae), Trplochiion scieroxylon (S tercu] iaceae), Buinesia spp., Guiacuin spp., Zygophyllurn
spp. (Zygophyllaceae).
Parenchyma strand = a series of axial parenchyrna cells forrned through transverse divi-
sion(s) of a single fusiform cambial initial cd.
Two cells per parenchyma strand, figs. 103, 104, e.g., Dalhergia spp., Lonchocarpus
spp., Pterocarpu.s spp.(Papilionaceae),
Four (3-4) celis per parenchyma strand, fig. 104, e.g., Termina/ia spp. (Combreta-
ceae), Ligusrtim spp., Syri;iga spp. (Oleaceae), Nesogordonia spp. (Sterculiaceae).
Eight (5-8) edis per parench yma strand, fig. 105, e.g., Nerium oleander (Apocy-
naceae), Macaranga spp. (Euphorbiaceae), Fraxinus spp. (Oleaceae).
Over eight celis per parenchyma strand e.g., Bhesa spp. (Celastraceae), Lophira
spp. (Ochnaceae), Minquartia spp., Tetrasylidum spp. (Olacaceae).

Commenrs.
Type ai' parenchyma, fusiform vs, sirand, is dctermined frorn tangential sections. Fusiforni
parenchyma celis are relatively uncomrnon and generaily occur in woods with storied structure
and short axial elements. In some species, cornbinations of the above fea[ures occur, e. g. fiisi-
fopn cdlls' and two cel]s per parenchyma sirand', or 'twa cel]s per parcnchyma srxand' and
'four (3-4) celis per parenchyma strand'. Strand length can difíerbetwcen earlywood and late-
wood of the sarne ring, ar between vessel-associated parcnchyma and parenchyrna which is not
in concact with the vesseis. Record ali cornrnonly occurring strand lengths.
Cawion.' Be careful not to confuse uniseriate rays or septate fibres with strand parenchyrna. Do
nor determine number of ce]ls per strand from chambere.d crystalliferous sirands.

95. Uniignifled parenchyma

Fig. 103. Fusiform parenchyma ce]Is (feature 90, left and right) and twa celis per parenchyrna
Definition: su'and (feature 91, left centre). Note also feature 120 (axial parenchyrna storied) and fcatiire 142
Uniignitied parenchyma = as per feature descriptor, fig. 106, e. g, Apeiba spp., Ente/ca (crystals in chambered axial parenchyina). Aeschynomcne e/aphroxylon, x 115. - Fig. 104.
arborescens, Heliocarpus spp. (Tiliaceae), Laporteo srimtdans (Urticaceae). Two (feanire 91) ia fom (feature 92) edis per parenchyma strand, Cordia abyssinica, x 115.
Fig, 105. Figiu (feature 93) ar over eight ceils (feature 94) per parenchyma strand, Bertholleíia
Conuneni: excelsa, < 115.—Fig. 106. Un]ignified parcnchyma (feature 95), Ente/ea arborescens, x 45.
Unlignified parenchyma usually occurs in broad bands, and is resu-icted to a small number aí
taxa.

282 IAWA Bulietin n.s., Vol. 10 (3), 1989 IAWA List of microscopic features for hardwood identification 283

RAYS

RAY WIDTH 1
J.
F^Q
96. Rays exclusively uniseriate
97. Ray width 1 to 3 celis
98. Larger rays commonly 4- to 10-seriate
99. Larger rays cornmonly > IO-seriate
100. Ra ys with multiseriate portion(s) as wide as uniseriate portions

Definitions:
Ray width in celi numbers as per feature descriptors.
Rays exclusively uniseriate, fig. 107, e. g, Lophoperalwn beccarianum (Celastraceae),
Termina/ia superba (Combretaceae), Hura crepitans (Euphorbiaceae), Castanea sariva (Faga-
ceae), Populus spp. (Salicaceae).
Ray width 1 to 3 celis, fig. 108, e.g., Aucoumea klaineana (Burseraceae), Dialium gui- 104
anense (Cacsalpiniaceae), Alseodaphne costa/is (Lauraceae), Albizia (Samanea) saman (Mimosa-
4Ij.' 1

ceae), Malus co,nmunis (Rosaceae).

Larger rays commonly 4- to 10-seriate, fig. 109, e.g., Acer saccharum (Aceraceae),
Spondias mombin (Anacardiaceae), Anisoprera laevis (Dipterocarpaceae), Khaya anthotheca
(Meliaceae), Celtis sinensis (Ulrnaceae).
Larger rays comrnonly > 10-seriate, fig. 110, e.g., Quercus spp. (Fagaceae), Fora-
queiba guianensis (Icacinaceae), Rapanea spp. Myrsinaceae), Platanus spp. (Platanaceac), Card-
we/lia sublimis, Grevi/lea robusta (Proteaceae), Tamarix aphyl/a (Tamaricaceae), Jacquinia revo-
luta (Theophrastaceae).
Rays with multiseriate portion(s) as wide as uniseriate portions, fig. 111, e.g.,
Anthodiscus amazúnicus, Caryocar cosraricense (Caryocaraceae), Srrombosja pustu/ata (Olaca-
ceae), Ad/na cordifolia (Rubiaceae), and many Apocynaceae, Sapotaceae, and Euphorbiaceae.

Procedure:
Determine ray width on the tangential section by counting the number of celis in the widest
part of the rays, perpendicular to the ray axis. When rays are of two distinct sizes (feature 103),
record the width of the larger sire class in the database.

Commen Is:
'Exclusivel y uniseriate rays' and 'rays> lO-seriate' are the Ieast comnion of the ray width
features. The categories for ray width match those of the Clarke (1938) and the Princes Risbor-
ough multiple entry keys (Brazier & Franklin 1961). There are some taxa that may be separated
by uinerdistinctions within these categories, and in adescription more detailed information than Fig. 107. Rays exclusivcly uniscriate (featiire 96), Lophopetalum beccarianuin, x 75. - Fig.
provided by these categories would be desirable. 108. Ray width one to three celis (feature 97), Alhizia (Samanea) sarnan, x 29. - Fig. 109.
Caution: These features for ray width do not apply tu rays containing radial canais (feature 130) Largcr rays commonly 4-10-seriate (feature 98). Note also shcath ecOs (arrow, feature 110);
Celtis sinen-sis, x 29. - Fig. 110. Larger rays commonly> lO-seriate (feature 99), Cardwellia
or to the rays composing an agegate ray (feature 101). Thus, for a wood with exclusively uni-
subi/mis, x 19. Fig. 111. Rays with miiltiseriate portion(s) as wide as uniseriate portions
senate rays, except for the rays with radial canais, describe the wood by recording feature 96 for (feature 100), Caryocar costaricense, x 115.
exclusively unisenate rays and feature 130 for radial canais.

284 JAWABulietin n. s., Vol. 10 (3), 1989 IAWA List of microscopic features for hardwood identification 285

AGGREGATE RAYS

101. Aggregate rays

DeJlnirion:

Aggregate ray = a number of individual rays so closely associated with one another that
they appear macroscopica]ly as a single large ray. The individual rays are separated by axial
elements, e.g., many species ofAlnus (Bctulaceae), Carpinus, Cory!us (Corylaceae), Casuarjna
(Casuarinaceae), Necepsia afzelii (Euphorbiaceae), Castanopsis, Lithocarpus, Quercus - ever-
green species (Fagaceae), Enimotum orbiculatum (Jcacinaceae), Cryptocarya densflora (Lauia-
ceae).

Commenrs:
There is variation in the size of the individual rays of aggregate rays. In some specics the ag-
gregate rays are composed of narrow rays (figs 112, 113, e. g., Carpinus spp. - Corylaceae),
while in others they are composed of broad rays (figs. 114, 115, e. g, Empnotum orbiculatu,n -
lcacinaceae).
Aggregate rays occur in few taxonomic groups.
Caution: Aggregate rays may be relatively infrequent inche taxa in which they occur, so they
may be easily overlookcd or absent in a small sample; therefore, this feature should preferably
be used posiive1y only.

RAY HEIGHT

102. Ray heiglit > 1 mm

Definition:
Ray height> 1 mm = the large rays commonly exceeding 1 mm in hcight, e.g., Guatteria
schomburgkiana (A nnonaceae), Anisoptera laevis (Dip terocarpaceae), Uapaca guineensis
(Euphorbiaceae), Scotzellja coriacea (Flacourtiaceae), Barrington(a asiatica (Lecythidaceae),
Pia-
ianus occidentalis (Platanaceae), Paypayroia guianensis (Violaceae).

Procedure:
Determine total ray height in tangential section, along the ray axis.

Comnzent:
In this Iist of features, only one category for total ray height is used as was done in some of
the earlier multiple entry keys (Clarke 1938; Brazier & Franklin 1961). More detailed ray height
data generally are given in descriptions and may be helpful in distinguishing between taxa in
some groups. Ray height is quite variable in some woods (particularly woods with markedly
heteroceilular rays), bui quite uniform in others (particularly woods with storied strucwre).
Figs. 112-115. Aggregate rays (feature 101). - 112 & 113: Carpinus herulus, aggregatc r,r:
(ar) composed of narrow rays. - 112: Transverse section, x 20. - 113: Tangencial sectr.
x 52. - 114 & 115: E,nmozum orbicularum. aggregate ravs (ar) composcd of rnu1tisriate
45. - 114: Transverse section. - 115: Tangential seclion.

286 lAWABufletjn n.s., Vol. 10 (3), 1989 1 AWA List of microscopiC features for hardwood identitication 287

RAYS OF T\VO DISTINCT SIZES

,
103. Rays of two distinct szes -, E f

Definition:

Rays of two distinct sizes = when viewed in tangential section, rays form two distinct
populations by their width and usually also by thcir height, figs. 116, 117, e.g., Acer saccha-
rum (Aceraceae), Poga oleosa (Anisophylleaccae), Ilex aquifolium
(Aquifoliaceae), Dillenia
pentagyna (Dillcniaceae), Quercus spp. (Fagaceae), Dendrobwgia boliviana
(Icacinaceae), Sca-
phiwn macropodum (Sterculiaceae), Ternsrroernja spp. (Theaceae). -k

Commenis:
There are no limits for the size classes - the smaller rays may be 1- or 2- or 3-seriate, the +
]arger rays may be less than 5-seriate.
Generaily, to fit the feature definition, intcrmecliate rays should not exist between the two
populations or be quite rate. Thus, when very large rays occur with a fcw rnedium-sized and
more numerous small rays (e. g, Fagu.$), feature 103 'rays of iwo distinct sizcs' may still be
applied.

Cautions: Vhether a wood has rays of two distinct widths cannot

be deterrnined from the cross 1 ..
section because in this view the long uniseriate wings of heteroceilular multiseriate rays might bc
interpreted incorrecdy as narrow rays.
Aggregate rays per se houId not be considercd as a separate ray size class. Only iri those Figs. 116 & 117. Rays of two distinct sizes (feature 103). - 116: Ternstroemia sp., x 45. - 117;
species where the aggrcgate rays are composed of much broader rays than thc nonaggregate rays Quercus gilva, x 29.
does feature 103 apply, e. g,, severa] species of Ca.suarina (Casuarinaceae) and Quercus (Faga-
ceae).
 288
IAWA Bulietin n.s., Vol, 10 (3), 1989 IAWA List of microscopic featurc.s for hardwood identilication 28')

RAYS: CELLULAR COMPOSmON

104. Ali ray cells procumbent

105. Ali ray celis upright and/or square
106. Body ray ceiis procumbent with one row of upright and/or square niargi-
na! ceils
107. Body ray ceiis procunibentwjth moslly 2-4 rows of upright and/er square
marginal celis
108. Body ray celis procumbent with over 4 rows of upright and/or squarc mar-
ginal ceils
109. Rays with procumbent, square and upright celis mixed throughout tlie ray

Definjrfons:
Procumbent ray cefi = a ray parenchyrna ceil with its longest dirnension radial as seen in
radial section.
Square ray celi = a ray parenchynia ccli approximatel y
square as seen in radial section.
Upright ray celi = a ray parenchyrna cel] with its lorigest dirnension axial as seen in radial
section.

AR ray cetls procumbent, fig. 118, e. g., Acer spp. (Aceraceae),

Tabebuia spp. (Bigno-
niaceae), Alhizia spp. (Mimosaceae), Hannoa klaineana
(Sirnaroubaccae)
Ali ray celis upright and/or square, fig. 119, e.g., Hedyosmum scabrum
thaceae), Aucubajaponica (Cornaceae). (Chloran-

Body ray ceils procumbent with one row ol upright and/or square marginai
celis, fig. 120, e.g., Kalopanax pictus (Araliaceae),
Aucm.mea klaineana (Burseraccae),
Pseudocedrela kanschyi (Meiiaceae).

Body ray celis procumbent with mostiy 2-4 rows of upright and/or square
marginal celis, fig. 121, e.g., Liq uidainbarstyrac,Jlua
(Hamameljdaceae), Carapa guianensis
(Meiiaceac), Trecu lia africana (Moraceae), A/seis peruviana
(Rubiaceae), Euscaphf.s spp, (Sta-
phyleaceae).

Body ray eeils procumbent with over 4 rows of upright and/or square margi-
nal cells, fig. 122, e.g., Wein,nannia descendens (Cunoniaçeae),
Quintinia spp. (Escailortia-
ceae). HomaIiumfoetid (Flacourtiaceae), Hu.'niria spp,
(Humiriaceae), O?oschulzj spp.
(Icacinaceae), Coffea spp. (Rubiaceae), Turpinia spp. (Staphvleaceae).

Rays with procumben, square and upright celis mixed throughout the ray,
'T.
fig. 123, e. g, Guaerja spp. (Annonaccac),Xan,/jop1/u, lanceatum
(Polygalaceae), Pometia
pinna:a (Sapindaceae), Heliocarpus spp. (Tiliaceae). I1tlii t

Fig. 118. Ali ray edis procurnbent (feature 104), Acer cwnpcslre. x 115.-- Fig. 119. Ali ray
ecOs upright anrifor squarc (feature 105), Aucuba japonica, x 29.— Fig. 120. Body ra y cells
procumbeni with oric row of upright and/ar square marginal edis (feature 106). Fseudocedrelu
kotschyi, x 115.— Fig. 121. Body ray celis procumhent with most]y two to íoui' rows of up-
right and/or square marginal cells (íeature 107), Campa guianen.sis.. x 114. Fig. 122. Body
ray celis procurnbent with mostly over 4 rows of upright aiidlor squarc marginal celis (feature
108), Hornalium j.'etidwn. x 115. - Fig. 123. Rays with procumbeni, square and upright cclls
rnixed throughout thc mv (feature 1091, Xwtthophyl!wn lanceatwn, x 90.
 290
EAWA Builetin ris., Vol. 10 (3), 1989
Frocedure;
Use radial sections LO determine the cellular composition of rays because types of ray celi
(procurnbent, upright, and square) are defined on the basis of their appearance in radial section.
Generaily, upright and square cens, if presem in eomhinatjon with procumberiL celis, are located
ri
Lhe marginal rows, i.e,, those rows at lhe top and bottom of lhe ray, and procumbcnt cells are
]ocatcd in lhe body (centre) of lhe ray.
In woods wih uniseriate and rnultiscriate rays - desenhe lhe celhila.r cocnposition of l he mui-
tiseni ate rays, not lhe uniseriate rays. Some woods have more than onc category of ray type with
rcspect to celiular composition, only record lhe rclatively common categonies.
Cornmen13-
The ceiluiar colnposition of lhe niultiseriate and uniseniate rays in [f i e sarne wood is not
necessariiy the sarne. In some woods, their uniseniate rays are cornposed oniy of upright celis,
while thcir multiseniate rays are composed of both upright and procurabeni celis.
Hornoceilular rays are rays coniposed of a single ccli type; heteroce]lular rays are composed
of two ar more ccli types.
Thc tens homogencous and hetemgenenus are used to describe ray tíssue as a whole. De-
signating Krihs types (Kribs 1968) rnay be useful when describing a wood, but are not used in
this Iist. Roughly spealdng, Kribs homogeneous types correspond with feature 104 (ali ra y celis
procurnbent), Kribs heterogeneous 111with feature 106 (one row of upright and/ar .square mar-
ginal edis), Kribs heterogeneous II with fcature 107 (2-4 rows af upright and/or square mar-
ginal celis), Krihs heterogeneous 1 with feature 108 (more than 4 rows of upright and/or square
marginal celis). H&wever, feature 108 also may partly ovenlap witb Kribs heterogcneotjs 11
body of the rays is very high. II' ffie
IAWA List of microscopic features for hardwood identification 291
Cautions:
Ray composition often varies betwecn juvenile and mature wood. Ia many species, rays mar
the pith may be composed entirely of upnight celis, while rays distant from lhe pith are com-
posed largely of procumbent edis with only a few rows of upnight and/or square edis. When
creadng a database, only examine mature wood samples or, for shrubs, lhe penipheral wood of
lhe thickest available stems. When an unknown wood fragrnent is from a thin branch, do nor use
ray composition.
Although in tangential section, marginal rows of upright and/or square edis often will ap-
pear as uniseriate margins, lhe presence of uniseriate margins alone is not a reliable indicator of
heteroceliular rays. Ia some woods (e, g., Carya spp. —Juglandaceac), there are uniseniate mar-' -.
ginal rows visibie in tangential section, and these celis appear larger xhan the body edis, but
wheri viewed in radial section these edis are procumbent, as are lhe edis of the rnultiseriatc por-
tion.
Sheath celis (feature 110) or tile cells (feature 111) are not considered when deternilning ray
ceiluiar composition.
Feature 109 only appiies if there is a mixture ar alteration of different ray cdl shapes xhrough-
out the ray, irrespective of whether it is in uniseriate or nnildseniate rays or ray portions. Feature
109 does not appiy to woods with vertically fused rays (ar 'rays with alternating uniseriate and
multiseriate portions') where the uniseriate portions may be composed of square and upright
cells and lhe multiseniate portions ofprocumbent edis. Just lhe presence of sheath ce]is (feature
110) or tile edIs (feature lii) also does not qua]ify a wood for frature 109.
pr-

292 IAWA Builetin n.s., Vol. 10 (3), 1989 IAWA List of microscopic features for hardwood identitication 293

SHEATH CELLS

110. Sheath celis

DeJlnition:
Sheath celis = ray cells that are located along thc sides of broad rays (> 3-seriate) as view-
ed in tangential section and are larger (generaily tailer than broad) than the central ray celis, figs.
124, 125, e. g., Ceiba pentandra (Bombacaceae), Cordia altiodora (Boraginaceae), Sainbucus
nigra (Caprifoliaceae), Dipterocarpus lowii (Dipterocarpaceae), Stemonurus luzoniensis (Icaci-
naceae), Ailanthus altissima (Simaroubaceae), Sterculia oblonga (Sterculiaceae).

Commenrs:
Presence of sheath celis should be deterrnined from tangential sections. There is variability in
the frequency and distinctiveness of sheath celis. In some species most, if not ali, multiseriate
rays have sheath cells which are much larger than the other ray celis, while in others sheath celis

IiuI11
are not frequent and/or slightly larger than the adjacent cells. When identifying an unknown
wood sample, do not use this feature as a flrst lime of approach unless it is well mai-ked.
Cautwn: Do not confuse sheath celis with tile cells (feature 111), which are always found in the
body of the ray as weli as the edges and are visible in both tangential and radial sections.

TILE CELLS

111. Tile celis

Definition:
Tile celis = a special type of apparently empty upright (rarely square) ray celis occurring in
intermediate horizontal series usually interspersed among rhe procumbent celis, figs. 126-129,
e. g., Durio spp. Neesia aitissinia (Bombacaceae), Guazutna spp., Kleinhovia hospira, Prero-
spermum spp. (Sterculiaceae), Despiarsia spp., Moi/ia spp., and some species of Grewia (Tilia-
ceae).

Commenrs:
Tile edis sometimes have been classified into two groups: type Durio when they have the
sarne height as the procumbent ray celis, and type Pterospermum when they are higher. How-
,L, ''11fr1
ever, this distinction is dubious because there are intergradations between the two, as in Gua-
ruma (Sterculiaceae), and Grewia (Tiliaceae).
Tile edis do not occur in uniseriate rays, and as far as is known are restricted to the order
Malvales.

Figs. 124 & 125. Sheath celis (feature 110). - 124: Ceiba pentandra, x 75. - 125: Stemonuru
luzoniensis, x 75.— Figs. 126-129. Tile celis (feature 111). - 126-128: Durio type, Neesia
alrissima, radial, transverse, and tangential section, x 115. - 129: Pterospermum type, Ptero-
spermwn grewiaefoliurn, radial section, x 120.
Iii
PPP-

294 lAWABulietin n. s., Vol. 10 (3), 1989 IAWA List of rnicroscopic_features for hardwood identification 295

PERFORATED RAY CELLS

112. Perforated ray celis

Definition:
1
Perforated ray celis = ray celis of Lhe sarne dimensions or larger than the adjacent edis,
but with perforations, which generaily are on the side walis connecting Iwo vesseis on either
side of the ray, figs. 130-132, e. g., Combretwn Ieprosrachiuni (Combretaceae), Richeria race-
mosa (Euphorbiaceae), Chaunochiton breviflorum (Oiacaceae).

1
Commenrs:
The type of perforation in a perforated ray ccli may be simple, scalariform, reticulate, or for-
arninate, and does not necessarily coincide with the type of perforation plate occurring in the
vessel elements of the sarne wood. For instance, Sloanea monosperma (Elaeocarpaceae) and
Richeria racemosa (Euphorbiaceae) have simpie perforations in the vessei elernents and scalari-
form perforations in Lhe perforated ray cells. In Siparuna (Monimiaceae) there is a range of
multiple perforations in thc ray celis, but the vessel elemcnt perforations are simple and/or sca-
iariform with variations depending on species.
Pcrforated ray celis have bordcrcd pits similar to the intervessel pits. They can occur mdi-
vidually or in radial ortangential rows. Radial rows ofperforated ray edis with perforations in
the tanendai walls have been described as radial vesseis (Van Vliet 1976a).
Caution: Use this feature positively only and with some caution because spccies that have per-
forated ray celis may have thein in such low frequency that they could easily have bcen over-
iooked when creating a database, or exarnining an unknown.

DISJUNCTIVE RAY PARENCHYMA CELL WALLS

ft .
113. Disjunctive ray parenchyrna ccli walis
u

Definirion:

Disjunctive ray parenchyma ceil walis = ray parenchyma edis partialiy disjoined but
with contacts maintained through tubular or compiex wall processes, fig. 133, e. g., Funtuinia
africana (Apocynaceae), Buxus sempervirens (Buxaceae), Crozon oligandrus, Glycydendron
a,nazonicum, Suregada laurina (Euphorbiaceae), Malpighia incana (Malpighiaceae), Gardenia
imperialis, Randia armara (Rubiaceae).
135
Comrnent:
Axial parenchyma nay also be disjunctive
Figs. 130-132. Perforated ray celis (feature 112). - 130: Simple perforation in radial wai.I,
Chaunochiron breviflorum, x 115. - 131: Reticulate perforation in radial wali, Richeria race-
mora, x 115. - 132, Sirnple perforation in tangential wali, Com brerum leptosrachium, x 290. --
Fig. 133. Disjunctive ray parcnchyma ccli walis (feature 113). Malpighia incana, x 290.
Figs. 134 & 135. Wood rayless (feature 117), t'eronica tr2erSii, x 115. 134: Transverse sec-
tion. 135: Tangential section.

296 lAWABulietin n.s, Vol. 10 (3), 1989 1 AWA List of rnicroscopic features for hardwood identification 297

RAYS PER MILLIMETRE W000 RAYLESS

114. 54/mm 117. Wood rayless

115. 4-12/mm
116. ^ 121 mm Definition:

Definirions:
As per feature descriptors, examples follow.
^ 4/ mm, e. g., Guarreria schomburgkiana (Annonaceae), Cussonia arborea (Araliaceae).
4-12/mm, e. g., Acer rubrum (Aceraceae), Acacia spp. (Mimosaceae).
^: 12/ mm, e. g., Diospyros ,nesp il(for?nis (Ebenaceae), Randia armara (Rubiaceae).
Procedure:
The number of rays per linear unit is best determincd frorn a tangential section along a une
perpendicular to the ray's axis; it can also be deterrnined from a cross section. Make ten meas-
urements and record the categories the range falis within.
Com ments:
The feature 'rays 4-12 per mm' is more common than the features 'rays ^. 4 per mm' or
12 per mm' (Metcalfe & Chalk 1950). The number of rays per mm cannot sensibly be de-
termined in woods with aggregate rays, or woods with very broad rays and mo distinct size
classes, e. g., Quercus spp. (Fagaceae).
Wood rayless = wood with only axial clements, figs. 134, 135, e.g., Arrhrocne,nu,n
,nacrostachyum (Chenopodiaceae), Heimerliodendron brunonianum (Nyctaginaceae), Hebe sah-
czjolia, Veronica rraversii (Scrophulariaceae).
Commenr:
Rayless woods are restricted to a small number of families (Carlquist 1988).
Caurion: In rayless woods with included phloem (e.g., several Chenopodiaccac. the conjunc-
tive parenchyma (i.e., the parenchyma linking mo or more phloem strands) may forni radiai
extensions which resembie rays. In such woods there rnav be a continuum from short radi1
wedges tu long radial strips to 'normal' rnuitiseriai ravs lhn r al. ) tl teLLue fd\
1es' should he iised i[li caution in such 'oods
298 IAWA Builetin as., Vol. 10 (3), 1989 IAWA List of rnicroscopic features for hardwood idendfication - - 299

118. Ali rays storied

119. Low i-ays storied, high rays nonstoriect
120. Axial parenchyma and/or vessei elemcrits storied
121. Fibres storied
122. Rays andlor axial elements irregularly storied
123. Number of ray tier: par axial mm

Definitions:
Storied structure = cclls arranged in tiers (horizontal series) as viewed on lhe tangenrial
surface.
Ali rays storied, 1lgs. 136, 138, e. g, Dalbergia bariensis, Prerocarpus sansalinoides (Pa-
piiionaceae), Quassia amara (Simaroubaceae).
Low rays storied, high rays nonstoried, fig. 137, e.g., Scaphium spp., Triplochiton
scleroxylon (S tercul iaceae), Cercis canadensis (Caesalpiniaceae).
Axial parenchyma and/or vessel elements storied, fig. 136, e.g, Balanites aegyp-
daca (Balanitaceae), Dalbergia bariensis, Sparziwnjunceum (Papilionaceae), Tarnarix spp. (Ta-
maricaceae).
Fibras storied, 6g. 138, e. g, Quassia amara (Simaroubaceae), Ocwmeles sumacrana (Da-
tiscaceae), Zygophyllu.in dumosum (Zygophyllaeeae).
Rays and/or axial elements irregularly storied = stories of rays and/or axial ele-
ments not horizontal or straight, but wavy or oblique (synonym: in echelon), or only locally
presem, fig. 139,e.g., certain Leguininosae (Monopetalantlius, Teiraberlinia), Entandrophrag-
rna cylindricwn (Meliaceae), Fraxinus alba (Oleaceae), Tieghemella spp. (Sapotaceae).
Number oF ray tiers par axial mm = as per feature descriptor.

Co,nments:
lhe
The presence of storied structure should be deterrnined &om lhe tangential section, foi
radial section. These features can be recorded singly or ia combination as in some woods ali
elements are storied (e. g., Hibiscus tiliaceus -- Malvaccae, Cenrrolobiu,n paraense, Afrormosia
data - Papilionaceae), whiie in cahers various combinations of elernents are storied. Generally,
if parenchyma is storied, the vessel elements also are storicd. Featare 122, 'rays and/or axial
elements irregularly storied', is used in combination with lhe other features when appropriate.
There is variability within species and saniples. For instance, ia some samples ofSwietenia
(Meliaccae) rays are defiriitely storied, in others irregulariy storied, and in still others rays are
aol storied.
Tiers of rays are visibie at low magnification, or with lhe unaided eye or a hand lens, and
appear as fina horizontal striations or ripple marks on the tangencial surface. Especiafly ia lhe
Leguminosae, which has many taxa with storied rays, the number of ray tiers per mm (feature Fig. 136. Ali rays storied (fearure 118) and axial parenchyrna and/ar vessel elements storied
123) can be useful in distinguishing genera and specics. (feature 120), Dalhergia bariensi.s, x 75. Fig. 137. Low rays storied, high rays nonstoried
Absence of storied snuclure (features 118-123 abseni) is also oU diagnostic vaJue. (feature 119), Triplochiton scleroxylon. Note also features 120 (parenchyma storied) and 121
(fibres storied); x 45. - Fig. 138. Fibras storied (feature 121), Quassia amara, x 45. - Fig.
Caution: Storying of wide vessel elements may be obscured because oU ceil enlargement during 139. lrregularly storied structure (feature 122, hera expressed as rays ia echeion), Entandro-
vessel development, and se it is best to examine lhe narrower vessel elements te determine phragma cylindricwn, :x 45
whether vcssel elements are storied.

300 IAWA Bulietin n. s., Vol. 10(3), 1989
IAWA List of microscopic features for hardwood identification 301

SECRETORY ELEMENTS AND CAMBIAL VARIANTS

OIL AND MUCILAGE CELLS

124. Oil and/or mucilage celis aSsociated with ray parenchyma

125. 011 and/or mucilage edis associated with axial parenchyma
126. Oil and/or mucilage celis present anlong fibres

Definitions:
011 celi = a parenchymatous idioblast fihled with ou; mostly, but not always, enlarged and
rounded in outhne, occasiona]ly of considerable axial extension, figs. 140, 141, e.g., Nectan- ii .1ttIiI Ij
dra grandis, Ocotea glaucinia, O. tendia, Phoebe porosa and many other species of Lauraceae, mtt
Talaur,ia spp. (Magnoliaceae),
Mucilage ccli = a parenchymatous idiob]ast fihlcd with rnucilage; typically enlarged and
!
rounded in outline, occasionally of considerab]e axial extension (resernbling fibres), e.g., figs.
142, 143, axial iii some specics of Endlicheria (Lauraceae), and in ray parenchyma of some spe-
cies of Persea (Lauraceae).

Co,nments
Both oil cells and mucilage celis are commonly associated with axial and/or ray parenchynia, c
but may also occur among fibres. They are limired to very few woody dicotyledons and are
11J II
similar to one another, except for iheir contents, which are easily removed during rnicrotechnical
procedures (Richter 1977). 1.
Because ii is not practical to distinguish beween oil and mucilage edis by thcir appearance,
they are listed together. Various combinations of these features eccur togciher (see Baas &
1
Gregory 1985; Gregory & Baas 1989),

4 'j

Fig. 140. Oil and/or mucilage cdlis associated with ray pare nchyma (feature 124), Ocoiea giau-
cinia, x 115.— Fig. 141. Oil and/or mucilage edis associated with axial parenchyma (feature
125, arrows), Ocotea glaucínfa. Note also feature 124, ujl and/or mucilage edis in rays; x 45,
- - Figs. 142 & 143. Oil and/or mucilage cells presem among the fibres (feature 126), Ocorea
ccneilci (elongated inucilage celis), x 115.– 142: Transverse secdon. - 143: Radial section.

302 IAWA Builetin n.s., Vol. 10 (3), 1989 IAWA List of microscopic features for hardwood identiflcation 303

1NT1TRCELLI5LAR CANALS

127. Axial canais in long tangential lines

128. Axial canais in short tangential lines
129. Axial canais diffuse
130. Radial canais
131. Intercellular canais of traumatic origin

D/initions,
Interceilular canal = a tubular interceilular duct surroundcd by an epitheiium, gencrally
containing secondary piam products such as resins, gums, etc., secretcd by the epitheliai celis.
Interceiluiar canais may be oriented axialiy (axial/vertical interceliular canal), or radially (radial/
horizontal interceilular cana], within a ray), figs. 144-151. Synonyms: gum duct, resin duct.
Axial canais in long tangential iines = more than five canais in a une, fig. 145, e.g.,
Copaifera spp., Sindora spp. (Caesaipiniaceae), Dryobalanops spp, flopea spp., Neohalano-
carpus spp., Parashorea spp., J-'enracme spp., Shorea spp. (Dipterocarpaceae). Synonym: con-
cenic axial canais.
Axial canais in short tangential lines = two to five axial canais in a une, fig. 146, e. g.,
Diprerocarpus spp. (Dipterocarpaceae).
Axial canais diffuse = randorniy distributed soiitary canais, fig. 147, e. g., Prioria copai-
fera (Caesaipiniaceae), Anisoptera spp., Cotylelohium spp., Upuna spp., Vaiaria macrocarpa,
Varica spp. (Dipterocarpaceac).
Radial canais = canais presem in rays, figs. 148-150, e.g., Pistacia spp., Tapirira guia-
nen.sis (Anacardiaceae), Bursera gummzfera (Burseraceae), Shorea section Richetia (Dipterocar-
paceae), and rnany other genera of Anacardiaceae and Burseraceae. Synon ym: horizontal canais.
Trauniatic canais = canais formed in response to injury, arranged in tangential bands,
Li
generaiiy irregular in outline and ciosely spaced, fig. 151, e. g., Terminaliaprocera (Combrcta- ,_•,
ceae), Liquidambar sryracflua (Harnamciidaceae), Carapa procera (Meliaceae), Prunus serorinu '
(Rosaceae), Balfourodendrorz riedelianwn, Murraya exolica (Rutaceae), Quassia amara (Sima-
roubaceae). •

Cornments:
It is possibie to have a mixture of thesc features in one wood.
In some species of Dipterocarpaceae, the size of the axial canais is useful in differentiating
species, i.e., whether the canais are small (diameter < 100 xm) or largc (diameter> 100 im). FI
$:
The colour of resins in canais of Dipterocarpaceae can also be usefui in identif-ication.

1 z
: nu;
The effect of radial canais on ray shape (i.e., whether the canal niakes the ray fusiform in
shape or not), size, and number of canais per ray are also usefui features.
Caurion: Traumatic canais may not occur consistently in a species, therefore, when identifying
an unknown, never use the absence of traurnatic canais.
Fig. 144. Axial interceliular canal, Parashürea srn)rhiesii, x 290,— Fig. 145. Axial intcrcellular
canais in iong tangential hnes (fcature 127), Shorea parsfo1ia, x 29. - Fig. 146. Axial inter-
ceiluiar canais in short tangential lines (feature 128), Dipterocarpus grandiflorus, x 20. -- Fig.
147. Axial canais diffuse (feature 129), Vareriamacrocarpa, x 29.
1

304 IAWA J3ul]tin n. s., Vol. 10 (3), 1989 IAWA List of rnicroscopic feawres for hardwoud identification 305

II Itf .

¶1
l
v
149
,. ol I ' ^1,^

1
ÇJ .
!
i 151 ç ,

Figs. 148, 149. Radial canais (feature 130). 148: Bursera gurnmifera, x 70: - 149: Shorea h- Fig. 150. Radial canais (feature 130), Paraçhorea smythiesu. x 75. - Fig. 151. Jnterce11t11r
pefo1ia (arrow), x 140. canais of u-aurnatic origin (feanire 131), Murraya e.xotica, x 115-
 IAWA List of microscopie features for hardwood identification 307
306 IAWA Bulietin n.s., Vol. 10(3), 1959

TUBES / TUBULES

132. Laticifers or lanniniíerous tubcs

Dejlnitio:
Tubes/tubules = ce].]s or series of cclls of indeterniinate lcngth, extending radially rir ver-
iically (among fibres); according to specific contents mo Lypes can be distinguished.
Laticifers = tubes containing latex, the latex may be colourless or light yellow to hrown;
laticiíers rnay extend either radially (in geriera of Apocynaceae, Asclepiadaceae, Campanulaceae,
Caricaceae, Euphorbiaccae, Moraceae) or axial]y (interspersed aioong fibres and so Lar known
ori]y from Moraceae), figs. 152-155. Synonyms: latex tubes, latex canaIs.
Tanniniferous tubes = tubes containing tannins, whiçh are reddish-hrown, in rays (so Lar
known only froni Myristicaceae), Figa. 156, 157.

Corn,nenrs:
Although latex is often light-coloured, and tannins are dark, colour is not a reliable diffcr- 15
ence, and chenical tesis for tannin are needed to verify tube contents. Structural differences
between the laticifcrs and tarininiferous tubes appear minor (Fujii 1988). Thereforc, thesc mo
features are combiried into one descriptor. Latex traces are included in this clescriptor. It;
Tanniniferous tubes oftcn are difficult to recognise in tangcntial sections because in that vim
their dirnensions may appear similar to the ray edis; examining radial sections shows tannini-
ferous tubes to be longer than ray celis.

Figs. 152-157. Laticifers (latex tubes) or tarininiferous tubes (feature 132). - 152: Alsionia
scholaris, laticifer, x 140. - 153: Dyera costulata, wide laticiíers, x 180. - 154: Crotorr pana-
rnensis, laticifer with dark-staining contents, 290. - 155: Artocarpus conjnwjis, axial laticifcr.
x 75. - 156 & 157: Horsfieldia subg/ohosa, tanniniferous tube in transverse and radial seclion
(arrows), x 290.

308 IAWA Builetin n.s., Vol. 10 (3), 1989 1 AWA List of microscopic_features for hardwood identification 309

CAMBIAL VARIANTS

133 lncluded phloem, concentric

134. Included phloem, diffuse
135. Other cambial variants

Definrions:
tncluded phloem, coricentric = phloem sirands in tangential bands alternating with
zones aí xylem aiicl/or conjunctive tissue, fig. 158, e.g., Avicennia spp (Avic.enniaceac,
Suaeda monoica (Chenopodiaceae).
Incluided phloern, dilTuse = scattered, isolated phloeni strands. The phloem strands mav
be surrounded by parenchyma ar imperforate tiacheary elements, fig. 159, e,
vomica (Loganiaceae). Synonym: included phloem, foraminate ar island type.
Other cambial variants = caegory for a variety of cambial variants inclj:rr
cal, ílattened, and furrowed in cross section; axes with lobed ar furrowed xy1cn; isircc x:.
1cm; compound, divided, corded and cleft xylem masses.

Comrnents:
The features for includcd phloem type are based on the appearance aí [he wood, and do not
have devclopmental inferences - they are not defined on the basis of whether there is a sing]e
permanent ca'mbium, ar successive cambia, ar whethcr the tissue surrounding the phloeni Fig. 158. lncluded phloern, concentric (feattire 133, arrows), Suaeda ,nonoica. Note also radial
swands is xylern or conjunetive tissue. As pointcd oul by Mikesell and Popham (1976) and extensions of conjunclive parenchyrna (feature 117, wood rayless, variable-): x 38. Fig. 159.
Carlquist (1988), it is desirable to restrici ihe term interxylary ph]oem to chose cases in which the lncludcd phloern, diffuse (feature 134, arrows), Strychnos nux-votnica, x 45.
phloem has been produced internal]y by a single cambiurn.
lncluded phioem of the concentric type very often intergrades with diffuse included ph]oem
(e.g,, in many Chenopodiaceae); in all cases ofdoubt use both fcatures in species with concen-
tric included phloem the phloein bands may branch and anastomose, and the conjunctive parco-
chyma somedrnes fornis radial extensions resembling rays.
Because included phloem and other cambial variants are of regular occurrence iii the taxa in
which they are founcl, lhe cerrn 'anornalous' niust be considered a misnoiner.
Feature 135 'other cambial variants' most frcqucntly occurs in iiana: for more inforination
si±e Carlquist (1988),

310 -- - IAWA Builetin n. s., Vol. 10(3), 1989 1 AWA List of microscopic_features for hardwoodidentification 311

MINERAL INCLUSIONS

PRISMATIC CRYSTALS

136. Prismatic crystals present

137. Prismatic crystals in upright and/or square ray cells
138. Prismatic crystals in produmbent ray celis
139. Prismatie erystals in radial alignment in procumbent ray cells
140. Prisrnatic crystals in chambered upright and/or square ray celis
141. Prismalie crystals in nonchambered axial parenchyrna celis
142. Prismatic crystals in charnbered axial parenchyma cells
143. Prismatic crystals in fibres

Definirions:
Prisrnatic crystals = solitary rhombohedral or octahedral crystals coniposed of calcium
oxalate, which are birefringent under polarised light (fig. 160). Synonym: rhornboidal cxystal.
Chambered ceil = an axial parenchyma strand ceil or ray parenchyma ccli subdivided by
septa or b' thin to tliick celi walls.
1'
Features 137-143 as per descriptors, examples foilow. lf,,
Prismatic crystals in upright and/or square ray cells, fig. 161, e.g.. Astronium
spp. (Anacardiaceae), Bursera spp. (Burseraceae), Khaya anthorheca spp. (Meliaceae), 1-lelico-
stylis spp. (Moraceae).
Prismatic crystals in procumbent ray cells, fig. 162, e.g., Anogeissus laufolia (Com-
bretaceae), Carpinus spp. (Corylaceae).
Prisrnatic crystats in radial aligument in procumhent ray celis, fig. 162, e. g.,
Aspidosperma quebracho-bianco (Apocynaceae), A nogeissus bufo lia, Bucida buceras (Com-
bretaceae), Secw-i nega perrieri (Euphorbiaceae), Ouratea su.rinarnensis (Ochnaceae), Gonystylus
spp. (Thymclaeaceac). :.1FJ\
Prismatie crystals in chambered upright andfor square ray edIs, fig. 163, e.g..
Elaeocarpus cabomala (Elaeocarpaceae), Glycydena'ron a,nazonicum (Euphorbiaceae), Banara ni-
tida (Flacourtiaceae), Byrsonima laevigara (Malpighiaceae), Fagara fiava (Rutaçeae).

Fig. 160. Prismatic crystals (feature 136), Drypetes keyensis, x 230.— Fig. 161. Prismatic
crystals in upright and/or square ray edIs (feature 137), Astronium graveolens, x 115.— Fig.
162. Prismatic crystals in procumbent ray celis (feature 138) and in radial alignmcnt (fe.ature
139), Anogeissus laufolia, x 75. - Fig. 163. Prismatic crystals in chambered (or divided) up-
right and/or square ray cells (fearure 140), Elaeocarpus cabornala, x 290. - Fig. 164. Prísrnatic
crystals in nonchambered axial parenchyma edis (feature 141), Drypetes gerrardii, x 290. -
Figs. 165-167. Prismatic crystals in chambered axial parenchyma celis (feature 142). - 165:
Crysrals in short chains, Lithocarpus edulis, x 290. - 166 & 167. Crystals in long chains. -
166: Tangential section, Parkiapenduln,x 115, –167: Radial section, tvfalpighia incana, x 115.
F. 1 6. ['nsmatic crvstals in flhrcs (feature 14) Ranara reja.
115. .'

I1

312 IAWA Bulietin n. s -., Vol. 10(3), 1989 IAWA List of microscopic features for hardwood identification 313

Prismatic crystals in nonchambered axial parenchyma cells, fig. 164, e.g., Cei- DRUSES
ba spp., Ocliroma spp. (Bombacaceae), Drypetes gerrardii (Euphorbiaceae), Ficu..s spp. (Mora-
ceae). 144. Druses present
145. Druses in ray parenchyma cells
Prismatic crystals in charnbered axial parenchyma celis, figs. 165-167, e.g., Gil- 146. Druses in axial parenchyma celis
hertiodendron preusii (Caesalpíniaceae), Pentacrne contorta (Dipterocarpaceae), Lilhocarpus edu- 147. Druses in fibres
lis (Fagaceae), Juglans nigra (Juglandaceae), Couratari spp. (Lecythidaceae), Malpighia incana 148. Druses in chambered cells
(Malpighiaceae), Parkia pendula (Mimosaceae), Zanthoxylum hygrophila (Rutaceae), Manilkizra
spp. (Sapotaceae). Dejlnirions:
Prismatic crystals in fibres, fig. 168, e.g., Heniandradenia chevalieri (Connaraceae), Druse = a compound crystal, more or less spherical in shape, in which the many compo-
Banara regia (Flacourtiaceae), Triplaris americana (Polygonaceae), Punica granatum (Punica- nent crystals protrude from Lhe surface giving the whoie structure a star-shaped appearance,
ceae), Majidea zanguebarica (Sapindaceae). figs. 169-17 1, e. g., Hibiscus tiliaceus (Malvaceae). Syrionym: cluster crystal.

Co,nments: Features 145-148 as per feature descnptor, exampies foliow.

Prismatic (rhomboidal) crystals are the most common type of ciystal in wood (Chattaway Druses in ray parenchyma edIs, fig. 169, e.g., Gledirsia triacanthos (Caesalpinia-
1955, 1956). The relative abundance of prismatie crystals is variabie. In some species, crystals ceae), Macaranga barreri, M. heudelotrii Euphorbiaceae), Colubrinaferruginea Rhamnaceae),
are consistently abundam; in others, they are consistently prcsent, but not abundant; and in still Arnygdalus cammunis (Rosaceae), Celtis paniculata (Ulmaceae).
other species, they are present in some samples, but absent in other samples. In some taxa, crys-
tais occur in only one ceil type; in other taxa, they occur in more than orle ccli type. In the iatter Druses in axial parenchyma cells, fig. 170, e.g., Dacryodes edulis (Biirseraceac),
case, record ali features that apply. Terminalia catappa (Combretaceae).
Information on the specific location of crystals often is not available from the literature. Con- Druses in flbres, e. g., Combretum fruticosum (Combretaceae).
sequentiy, fature 136, 'prismatic crystals present' is useful in constructing a database from
existing information. Feature 136 is recorded in combination with other applicable crystal fea- Druses in chnmhered cells, fig. 171, e.g., Macaranga barteri, M. occidenralis (Euphor-
tures (137-143). biaceae), Banara regia (Flacourtiaceae).
In some species, crystals occur throughout Lhe ray (for heteroceliular rays: features 137 and
138 both present). In other species with heteroceliular rays they are restricted to the marginal Comrnenrs:
rows of upright and/or square celis, upright and/or square cells in the body of the ray, or sheath Most of the existing rnultiple entry ke ys and the literature do not provide information on the
celis (feature 137 present, feature 138 absent). In yet others, crystals are restricted to thc pro- location of druses. Consequently, the feature 'druses preseni' is included so that this informa-
cumbent body ray celis (feature 137 absent, feature 138 present). This last condition is not com- tion can be used.
mon and usually occurs in combination with crystals in a radial alignment (feature 139). The lat- In a given species, druses may occur in one or more of Lhe ccli types, and the celis may be
ter feature may appiy tu ciystals m nonchambered as weli as chambered celis. cnlarged as well (feature 156).
l'he descriptors 'prismatic crystais in chambered upright and/or square ray edis' (feature
140) and 'prismatic crystais in chambered axial parenchyma celis' (feature 142) include a con-
siderabie diversity in types of chambcred or subdivided celis (cf. Parameswaran & Richter OTHER CRYSTAL TYPES
1984) and, particularly for axial parenchyma, in length of the chains of crystalliferous chambers
or subdivisions. In some taxa there are oniy a few charnbers in a series, in others there are long 149. Raphides
chains. Such inforrnation should be recorded in a description. 150. Acicular crystals
151. Styloids and/or elongate crystals
Cauzions: There are many genera in which prismatic crystals are regularly absent (e. g,, Diptero- 152. Crystals of other shapes (mostiv small)
cai-pus spp. - Dipterocarpaceae, Betula spp. - Betuiaceae, Liriodendron Spp. - Magnoliaceae, 153. Crystal sand
and Ti/ia spp. - Tiiaceae). But, when identifying an unknown, using absence of crystals is no[
recommended because crystals are of sporadic occurrence in many other taxa (e. g., Acer spp. - DeJinition.',.-
Aceraceae, Quercus spp. - Fagaceae, and Ulmus - Ulrnaceae).
Care is needed to distinguish between crystals in septate fibres and crystals in charribered Raphides = a bundie of long needle-likc erystais, fig. 172, e. g .. Dilienia rericulara. Terra
axial parenchyma celis. cera boliviana (Diileniaceae), Pisonia Spp. (Nyctaginaceae), Psyc/lot ria recordiana (Ruhiaccac),
Tetrarnerista crassfolia (Tcuamcristaccac). Viris vinife ra (Vitaccae).
314 IAWA Bulietin n.s., Vol. 10 (3), 1989 IAWA List of microscopic features for hardwood identification 315

Acicular crystals = small needle-like crystals, not occurnng in bundles, fig. 173, e.g.,
Tecorna stans (Bignoniaceae), Cryprocarya glaucescens (Lauraceae), and Gmelina arborea (Ver-
benaceae).

Styloids = large erystais at least four times as long as broad with pointed or square ends,
fig. 174, e. g., Mayrenus obrusifolia (Celastraceae), Terminalia amazonica (Combretaceae), Gel-
semium senpervirens (Loganiaceae), Me,necylon membranfilium (Melastomataceae), Ga/lesia
inregrfo1ia (Phytolaccaceae), Gonystylus bancanus (Thymelaeaceae).

Elongate crystals = crystals two to four times as long as broad with pointed ends, fie.
175, e g., Siphonodon pendulwn (Celastraceae), Ligustrum vu/gare (Oleaceae), Vires glabratá
(Verbenaceae).

Crystals of other shapes (mostly smaH) = includes ali other shapes of crystals, figs.
176, 177, e. g., cubic (e. g,, Aporusa viliosa - Euphorbiaceae), navicular (boat.. shaped) (e. g.,
Litsea rericulara - Lauraceae), spindie-shaped (e. g., Dehaasia spp. - Lauraceae), pyramidal
(e. g., Caryodaphnopsis ronkinensis - Lauraceae), tabular (e. g., Aniba spp. - Lauraceae), in-
dented (e. g., Forestiera segregaza — Oleaceae), twinncd (e. g., Nesregis spp. —Oleaceae), etc.
-
i Crystai sarid = a granular mass cornposed of very small crystals, fig. 178, e. g., Cordia
suhcordata (B oraginaceae), Acrinodaphne hookeri (Lauraceae), Bumelia obtusifolia (Sapota-
.- ''... ceae), and Nico riana cordzfolia (Solanaceae). Synonym: microcrystais.
1 73

Com,nents:
Crystals, particularly the small ones, are best detected with polarised light

L: These crystals are not comrnon, and their occurrence may be sporadic. Therefore, these fea-
tures should only be used in the positive sense. Raphides and styloids often occur in enlarged
celis, feature 156. For more information on crystal types. see Chattaway (1955, 1956) and

- IL'
Richter (1980).

f
Caurions: Care must be taken not to interpret a cross section of an elongatc or acicular crystal as
a cuhic crystal.
lhe crystals in raphide hundies often scparate during sectioning.

lik OTHER DIAGNOSTIC CRYSTAL FEATURES

154. More than one crystal of about Lhe sarne size per celi or chamber
174 155. Two distinct sizes of crystals per ccli or chamber
156. Crystals in enlarged celis
157. Crystais in tyloses
Fig. 169. Druses (feature 144) in ray parenchyrna celis (feature 145), Hibiscus riliaceus, x 145. 158. Cystoliths
- Fig 170. Druses in axial parenchyma cells (features 144 and 146), Terminalia carappa,
x 170. - Fig 171. Druses in chambered ray celis (features 144, 145 and 148), Banara regia, Definir jons:
x 290. - Fig. 172. Raphides (feature 149) in procumbent ray ceil, Vitis vinifera, x 290. -
Fig. 173. Acicular or needle-shaped crystals (feature 150), Gmelina arborea, x 675. - Fig. Features 154-157 as per feature descriptor, examples follow.
174. Styloids (feature 151), Mernecylon memhrarnifolium (in included phloem), x 95. - Fig.
More than one crystal of about Lhe sarne size per celi or chamber, figs. 179,
175. Elongate crystals (feature 151), Ligustrum vulgare, x 220. - Fig. 176. SmaiI cubic crys-
182, e. g., Bouea opposirifolia (Anacardiaceae), Garcinia latissima (Guttiferae), Aniba dzsckeii
tais (feature 152), Aporu.sa villosa (in ray celis), x 290. - Fig. 177. SniaII navicular crystals
(f':iture 152), T,itsca reticulata. x 675. (Lauraceae). Ligustrum vulgare (Oleaceae), Gmelina arhorea. Vires divaricara (Verbenaceae).
316 IAWA Bulietin n.s., Vol. 10 (3), 1989 IAWA List of_microscopie features for hardwood identification - 317

Two distinct sizes of crystals per ceil or chamber, fig. 181, e. g_ Mangifera altis-
sima (Anacardiaceae), Cordia bantamensis (Boraginaceae), Pentacme conrorta (Dipterocarpa-
ceae), Zanthoxylum juniperinum (Rutaceae), Bumelia glomerara (Sapotaceae), Gonyszylu.s ban-
canus (Thymeiaeaceae).
Crystals in enlarged celis (idiobiasts), fig. 183, e.g., Carpinus carolinianum (Coryla-
ceae), Juglans nigra (Juglandaceae), Pyrts communis (Rosaceae), Citru.s auranrium (Rutaceae),
Carne/lia japonica (Theaceae), Zelkova serrata (Ulmaceae).
Crystals in tyloses, fig. 184, e. g., Astronium graveolens (Anacardiceae), Cordia gharaf
(Boraginaceae), Pera bumeliaefolia (Euphorbiaceae), Chlorophora tincroria, Pseudolmedia spu'-
ria (Moraceae), Chrysophyllum aurarum (Sapotaceae).
Cystoliths = internal stalked outgrowths of the ccli wali that project into the ccii lumen and
are composed of celiulose impregnated with calcium carbonate. They are irregular in shape and
sometirnes cornpletely 611 a ccli, fig. 185, e. g., in some Trichanrhera spp. (Acanthaceae), Spa-
rarra,uhelium spp. (Hernandiacae), and Opiiiaceae.

Comments:
Generaily, there is ordy one crystal per ccli or charnber. However, two or more similar-sized
crystals, especially acicuiar and/or navicular, and cubic and/or rectangular crystals, may Occur
in the sarne ccli or chamber. li is rare that there are two distinct sizes of crystals in the sarne ccli
or chamber. For more information, see Chartaway (1956).
For feature 156, 'crystals in enlarged edis (idiohlasts)', the enlargcd edis can be either ray
or axial parenchyrna edis, or more rarely both. The crystals in enlarged edis may he prismatic
crystals, druses, raphides, or any other crystal type.
Cystoliths, as far as is known, occur only in the exampies given (Ter Weile 1980).

Caution: Raphides are bundies of crystals, but the whole bundie is considered as a singie unit,
and so feature 154, 'more than one crysta.l per ccli or chamber', does not apply to raphides.
Crystal sand should also not be coded under feature 154.

•; -

Iç
i. 1 Fig. 178. Crystal sand (feature 153, arrowheads), Cordia suhcordata (in ray celis), x 290.
Figs. 179 & 180. More than one erystal of about the sarne size per ccli or charnber (feature 154).
- 179: In axial parenchyrna celis, Garcinia lati ssima, x 75. - 180: In ray edis, Bouea opposiii-
folia, x 290.— Fig. 181. Two distinct sizes of crystals per cdl or chamber (feature 155), Cor-
dia banramensis (in ray celis), x 290. - Fig. 182. Different size classes of crystals intergrading
1
in some ray edis (feature 155 variable), but more than one crystai of about the sarne size per ccii
in others (feature 154 present), Cordia abyssinica, x 115. - Fig. 183. Crystais in enlarged celis
(featm-e 156), Citru.s aurantium, x 135. - Fig. 184. Crystals in tyloses (feature 157), Cordia
1
gharaf, x 140. - Fig. 185. Cystoliths (feature 158), Trichanthera gigantea, x 220.
 318 - - IAWA Bulietin fls., Vol. 10 (3),_1989
IAWA List of_microscopic_features for hardwood identificatjon - 319
SILICA

159. Silica bodies present

160. Silica bodies iii ray celis
161. Sulca bodies in axial parenchynia celts
162. Silica bodies in fibres
163. Vitreous silica

Defitjtjons:

Silica bodies = spheroidal or irregu]arly shaped particles composed of silicon dioxide.

Synonyms: sulca grains, silica inciusions.
Features 160 -162 as per feature descriptor, exampies follow.
Silica bodies in ray celis, fig. 186, e.g., in Tratrfnickia burserfolla, T. demararae (Bur-
seraceae), Licania leptostachya (Chrysobaianaceae), Shorea lameilara (Dipterocarpaceae), Mezi-
laurus irauba (Lauraceae), Vitex compressa (Verbcnaceae).
Silica bodies in axial parenchyma cells,fig. 187, e.g., inBomhax nervosum, Diste-
monanjhus spp. (Bomhacaceae), Apuleia leiocarpa, D ialium guianense (Caesalpiniaceae).
Silica bodies in fibres, fig. 188, e.g., Canariwn hirsutum, Prorium neglectum, Trarti-
nickia burserfaIia (Burseraceae), Ocotea pube ruia (Lauraceae).

Vitreous silica = silica that coats ccli walis or cornpletely fihis the ccli lurnina, fig. 189,
e.g., Srereospermum chelonoides (Bignoniaceae), Hydnocarpus graciüs (Flacourtiaceae), Arto-
carpus vriesianus (Moraceae), and Gynot-roches a.rlllw-is (Rhizophoraceae).

Procedures:
Sílica hodies: Silica bodies are observed with Lhe light microscope in radial sectioris of eithcr
permanent or temporary mounts or in celis that have been macerated, LI largo amounts of ex-
tractives are present and the sulca bodies are difficult to see in section, hleach with a domcstic
blcaching agent, rinse in water, heat in carbolic acid, and mount iri clove ou, or macerate a few
chips in any macerating fluid that removes most of the exlracives and lignin but not the silica,
At low magnification (4-10 x objective lens), si]ica bodies genera]ly appear as smaIl dark
nonbirefringcnt particles. At higher magnification (25-40 x objective lens), they have a 'giassy'
appearance.
Vitreou.s sílica.' Thoroughly macerate chips or splinrers, leave the wood iii the maceratirig
soludon until ii is white. Decant the macerating tluid, add water, rinse, decant, and repeat mui]
Lhe solution is clear. Place some macerated wood on a slide: warrn ilie slide on a hotplate until Fig. 186. Silica bodies present (feature 159) in ray edis (feature 160), Shorea lamellata, x 150.
- Fig. 187. Sulca bodies in marginal ray celis and axial parenchyma edIs (features 159, 160,
the macerated wood is dry. Aliow the slide to cool, and then add 2 to 3 drops of concentrated and 161), Apuleia lelocarpa, x 75. - Fig. 188. Silica bodies iii fibres (features 159 and 162),
su]furic acid to dissolve ihe ceilulose. Add a cover s]ip and observe the cel]s under a light micro- Ocotea ci. puberula. x 115. - Fig. 189. Viu-enus silica (feature 163),
scope ai low magnification. Vitreous silica appears like pieces of translucent vessel clements Hvdnocarpus gracilis,
x 5(}.
and fibres. To distinguish undissolved celis from vitreous silica use polarised light. Undisso]ved
edIs are bircfringent, whereas vitreous silica is not. Vitreous sulca can also be recognised in
weil-bleached sections bccause of irs 'glassy' appearance.

320

IAWA Bulietin n. s., Vol. 10(3), 1989 IAWA List of microscopic features for hardwood identification 321

Cornmenss:
APPENDIX
Silica bodies rnost often are rcstrictcd to ray celis, particularly the marginal or upright edis.
Non-anntomical Information
Sometimes they are rcsicted to axial parerichyma; somctimes thcy oceur in both ray and axial
parenchyma. Sílica bodies rarely occur in fibres, but if lhey do lhe fibres usually are septate.
Wheiher silica occurs in aggregations, as irregularly shaped or globular bodies, or whether GEOGRAPH1CAL DISTRIBUTION (fig. 190)
the silica bodies have a smooth or verrucose surface may be diagnostic in certain groups and
needs to be re.corded in a description. For more information on silica bodies, see Amos (1952), 164. Europe and temperale Asia (Brazier and Franklin region 74)
Ter Weile (1976), and Koeppen (1980). 165. Europe, excluding Mediterranean
166. Mediterranean including Northern Africa and Middle East
Cautions: When looking for silica bodies, do not use glycerin as a mounting medium because 167. Temperate Asia (China), Japan, USSR
its refractive index niakes it difficult to detect silica bodies. 168.. Ceniral South Asia (Brazier and Franidin rcgion 75)
Hydrofluoric acid, which is sometirries used to soften wood, wili dissolve the silica bodies. 169. India, Pakistan, Sri Lanka
170. Burma
171. Southeast Asia and the Pacific (Brazier and Franklin region 76)
172. Thailand, Laos, Vietnam, Cambodia (Indochina)
173, Indomalesia: Indonesia, Philippines, Malaysia, Brunei, Singapore, Papua New
Guinca, and Solomon Islands
174. Pacific Islands (including New Caledonia, Samoa, Hawaii, and Fiji)
175. Australia and New Zealand (Brazier and Franktin region 77)
176. Australia
177. New Zealarid
178. Tropical mainland Africa and adjacent islands (Brazier and Franklin regiori 78)
179. Tropical Africa
180. Madagascar & Mauritius, Runion & Comores
181. Southern Africa (south of the Tropic of Capricorn) (Ilrazier and Franklin rcgion 79)
182, North America, nonh of Mexico (Brazier and Frarddin rcgion 80)
183. Neotropies and temperate Brazil (Brazier and Franklin region 81)
184. Mexico and Central America
185 Carrihbean
186. Tropical South America
187. Southern Ilrazi]
188. Temperate South America including Argentina, Chile, Uruguay, and S. Paraguay (Brazier
--- - and Franklin region 82)
164
1B2<

IIs:fZ
Cü,nments.'
There is no single ideal way of dividing the world. Thc abovc is a mixture of political and
hiogeographica] criteria. It retains the major geographical regions of Brazie.r and Franklin
(1961), but some regions are subdivided so that more prccision is possible.

HABIT

189. Tree
190. Shruh
191. Vine/liana

Definjtion.ç:
Fig. 190, Geographical distribution (feaures 164-188).
Tree = woody perennial piam with one main stem usuatly over 3 Tnetres tal],
 R

322 IAWA Builetin n.s., Vol. 10 (3), 1989 IAWA List of microscopic features for hardwood identification 323

Shrub = a woody pererinial plant usual]y with several stems and usually Iess than 3 meti-es which specific gravity is detemiined. Basic specific gravity (Bsg), which is based on the green
uilI at maturity. volume (wood fully swollen, moisture coritent of fibre saturation point or higher) is one of lhe
Vine/liana = any plant witli a long relativel y thin siem that c!imhs along a suppnrt or trails most commonly cited values (Panshin & DeZeeuw 1980).
along lhe ground. Qffier values often given for wood include basic density (Bd) which is equal to lhe oven-diy
weight of woodfgreen volume and which has units (g/cm 3 , kg1m 3 ' or lbs/ft3 ). To convert
Comments: from basie specific gravity (Bsg) to basie density (Bd) rnultiply lhe basic s.g, by lhe density of
There wifl be overlap between these categorics in some species, and two or more, rarely ali water as is shown below:
three,may apply. Bd in gfcni 3 = Bsg x 1 glcm3
Bd in kg /M3 = Bsg x 1000 kg/m3
Bd in Ibs/ft 3 = Bsg ' 62.4 1b51ft3
WOOD OF COMMERCJAL HPORTANCE
Therefore
192. Wood of commercial importancc Bsg of 0.4 = Bd of 0.4 g/crn 3. 400 kg/m 3 , or 25 lbs/ft3
Bsg of 0.75 = Bd of 0.75 g/cm 3 , 75 kg1m3 , or 46.8 1b5/ft3
This category is intended for woods of both historical and current commercial importance.
The term ofcommereial importance' is somewhat vague, and should be used with caution when
identifying an unknown. But when identifying certain wooden artefacts, C. g., fiirniture. ii can
HEARTWOOD COLOUR
be helpful to segregate comrnercial species from noncommercial species.
196. Fleartwood colour darker than sapwood colour
197. Heartwood basically brown or shades af brown
SPECIFIC ORA vrry
198. I-teartwood basically red or shades of red
199. Ileartwood basically yellow or shades of yellow
193. Basic specific gravity 10w, ^ 0.40
200. Heartwood basically white to grey
194. Basic specific gravity medium, 0,40-0.75
201. Fleartwood with streaks
195. Basic speeilie gravity high, ^ 0.75
202. Heartwood not as above
Defini:ion.s:
Definitions:
Basic specifie gravity = ratio of lhe oven-dry weight of a piece of wood to the weight of Heartwood colour darker than sapwood colour, e.g,, Astronium spp. (Anaeardia-
lhe water displaced by lhe wood when it is completely swollcn (i.e., green volume). Examples ceae), Tabebuia guayacan (Bignoniaceae), Acocia koa (M.imosaceae), Morus alba (Moraceae),
for lhe feature categories follow. Robinia spp. (Papilionaceae).
Basic speciíic gravity Iow, :!^ 0.40, e. g., Dyera co.rtulata (Apocynaceae), Ceiba spp., Heartwood basically brown or shades of brown, e,g., Quercus alba (Fagaccae),
Qchrwn spp. (Bombacaceae), Populus balsamfera (Salicaceae). Triplochton spp. (Stercuiia- Albizia (Samanea) saman (Mimosaceae), Morus rubro (Moraceae), Eucalyptus giobulus (Myrta-
ceae), Ti/ia spp. (Tiliaceae). ceae), Robinia spp. (Papilionaceae).
Basie speciflc gravity medium, 0.40-0.75, e.g., Acer saccharum (Accraceae). Be- Heartwood basically red or shades of red, e. g., Brosimum rubescens (Moraceae),
tuia lenta (Benhlaceac), Carapa guianensis, Khava rardjf!iiia (MeiLeae). iruns aniercana Prerocarpus macrocarpus (Papilionaceae), Sickingia spp. (Rubiaceae).
(Oleaceae), Tectona spp. (Verbenaceae).
Ileartwood basically yellow or shades of yellow, e.g., Enantia chloranrha (Anno-
Basic speciíic gravity high. ^ 0.75, e. g ..As rrwuum 2reveo1ens (Anacard iaçc) rsaceae), Buxus spp. (Buxaceae), Schaefferiafrutescens (Celastraccae), Gossypiospernium spp.
Ocotea rodiei (Lauraeeae), Dalbergia melancwlon (Papilionaceae), Maniíkarn bidenrata (Sap) (Flacourtiaceae), Ciadrastis iutea (Papilionaceae).
taceae), Guaiacwn spp. (Zygophyl]aceae).
Heartwood basically white to grey, e.g., Jiex opaco (Aquifoliaceae), Didymopana.x
Com rnenrs: spp. (Araliaceae), Ceiba spp. (Bombacaeeae), Hura spp. (Euphorbiaceae), Cecropia spp. (Mo-
Density is the weight of a substance (mass) per uni[ volume: speciflc gravity (s.g.) is lhe raceae). Simarouba spp. (Simaroubaceae).
rario of the density of a material to lhe density of water and, conscquently, specific gravity does Heartwood with streaks, e. g., Dracontomelon dao (Anacardiaceae).
not have units. For purposes of compucing speci fie gruvily of wood, wood density uses lhe
oven-drv weight of wood as lhe numerator, Because lhe volume of wood changes with changes Heartwood not as above = colours such as black, purple, orange, green, as in Diospy-
in moisture content bclow fibre saturation point, it is neccssarv lo specify the muislure conreni ai ros ebenum (Ebenaceae), Peltogyne Spp. (Papilionaceae).
324 IAWA Bulietin n.s, Vol- 10 (3), 1989
Procedure:
Time and exposure to light may alter Lhe appearance or vividness of Lhe colour. Thercforc,
it is best to determine colour from a freshly cul tangentia] surface of a dry woocl sample. The
heartwood colour of recently felled trees often differs from that of dry wood samples. These
descriptors are for wood samples that are ai Ieast air-dry.
Comments.'
Feature 196, 'heartwood colour darkcr than sapwood colour', only can be used wben both
heartwood and sapwood are present, and is recorded in combination with Lhe other heartwood
colour features (197-202). The heartwood of some species, e.g., Quercus alba (Fagaceae) and
Betula lenta (Betulaceae), is not markcdly darker than Lhe sapwood, but can be distinguished
from it, so feature 196 applies to thcse taxa.
The variety of colours, shades, and combinations of heartwood colour make it impossible to
categorise ali of theni. In general, Lhe colour of heartwood is either brown, red, yellow, white,
or some shade or combination of these colours. Basically brown heartwood is very corumon;
basically red and basicafly yellow are rather rale; basicaily white or grey is rather frequent.
The heartwood colour of many taxa is not rcstricted to one colour, bui tu a combination of
colours and, when appropnate, various combinations shou]d be recorded and may be used when
idendfying an unknown.
Examples of these combinations include: reddish-brown in Astronium spp. (Anacardiaceae),
Hyrnenaea spp. (Caesalpiniaceae), Quercus rubra, Fagus spp. (Fagaceae), Khaya spp., Swiei-
enia spp. (Meliaceae); yellow and brown in Disternonanthus spp. (Caesalpiniaceae), Chloropho-
ra :inctoria (Moraceae), Adina cord(folta (Rubiaceae), Faga.ra spp. (I.utaceae), tvfastichodendron
spp. (Sapotaceae).
Very light coloured woods would be recorded as combinations of white to grey and brown
and/or yellow, c.g., Acer spp. (Aceraceae), Alstônia spp. (Apocynaceae), Anisoptera spp. (Dip-
terocarpaceae), Gmeilna spp. (Verbenaceae).
Heartwood with streaks is always uscd in combination wíth Lhe general heartwood colour, as
in Microberlinia spp. which lias brown heartwood with streaks.
'Heartwood not as above' is a 'catch-all' category for taxa with heartwood colours such as
black, green, orange, and purpie. This feature may be used alone (e. g., as for Diospyros ebe-
nwn - Ebenaceae, which has distinctly black heartwood), but more commonly it will be used in
combination with olher heartwood colours. For example, Lhe combinadon of basically brown
and green occurs in Bucida buceras (Combretaceac), Ocoteu rodiei (Lauraceae), Liriodendron
rulip (fera. Michelia spp., Talawna spp. (Magnoliaceae); Lhe combination of basically red, brown,
yellow, and orange with streaks occurs in Centrolobiu,n spp, (Papilionaceae) and Aspidosperma
spp. (Apocynaceae); lhe combination of brown, red, purple, black and orange with streaks oc-
curs in Dalbergia spp. (Papilionaceae).
Cautions:
Do not use heartwood colour features for ancient saniples, archaeological material, or other
samples whose colour has been altered by burial, time, lreatrnent, or decay.
Be particularly careful when using lhe feature 'Heartwood basically white io grey', because a
whitish coloured sample may represent sapwood and not heartwood.
IAWA List of mioscopic feanires for hardwood identi6cation -325
1
ODOUR*
203. Distinet odour
Definition:
Distinct odour = as par frature descriptor, e.g., Vihurnwn (Caprifoliaceae), Ceraropeta-
lum apetalum (Cunoniaceae), many specics ofLauraceae, Cedrela (Meliaceae), and Santalwn
(Santalaceae).
Procedure:
In dricd wood samples the chemicais responsible for Lhe odour rnay have voiatised froin lhe
surface, so it will be nccessary to expose a fresh surface, or take other measures to enhance Lhe
odour, e. g., add moisture by breathing on the wood, or wet Lhe wood with water and warm it.
Caucion: Odour is quite variable, and individual perceptions of odour often differ 'I'hercforc,
use this feature with caution and only ia Lhe positive sente.
* Taste is deibcrately excluded because of safety considerations, particularly a concern that
someone may try tasting a wood whose contents could cause a severe allergic reaclion.
HEARTWOOD FLUORESCENCE
204. Heartwood Iluorescent
Definition:
Heartwood fluorescent = heartwood fluorescing when ilhiminated with longwave ultra-
violei light, e.g., with a strong yellowish or greenish fluorescence in Anacardium excelsum
(Anacardiaceae), Asimina spp. (Annonaceae), Aspidosperina eburneum (Apocynaceae), Robinia
spp. (Papilionaceae); with a slight tinge of orange fluorescence ia Mangifera indica (Anacardia-
ccae), Varairen lundellii and Symphonia spp. (Guttiferae),Nauclea diderrichii (Rubiaceae); weak,
yet positiva, fluorescence in many Annonaccae, Lauraceae, and Magnoliaceae.
Procedure:
Samples for testing fluorescence must be frcshly surfaced; simply removing some shavings
with a knife is sufficient for exposing a fresh surface. Place samples i.mder a longwave (365 nm)
ultraviolet (UV) iight ata distance of less than 10 cm. A high-intensity longwave UV lamp and
observation in a darkened room is recommended.
Co,nrnents:
Fluorescent samples generally appear yellowish or greenish under Lhe UV iamp, although
some species show slight tinges of orange or pink.
Samples that are not fluorescent rnay rellect some of Lhe UV light making the sample appear
slightly blue or purpie. Some samples with a yellowish heartwood, such as Chloroxylon spp,
(Rutaceae) and Gonystylus spp. (Thymeiaeaceae), are not fluorescent, but may seem tu have a
weak yellow fluorescence because of reflection.
Absence of fluorescence can be irnportant in some famulies, e. g., Anacardiaceae and Legumi-
nosae. See Aveila et ai, (1989) for a survey of fluorescence ia Lhe dicotyledons.

326 IAWA Bulietin n.s., Vol. 10 (3), 1989
Caulions:
This feature applies only to naturaily occurrmg fluorescence and not to fluorescence asso-
ciated with decay or pathological infections. Wood infected with decay organisms may fluoresce
with sireaks, spots, or a mouled appearance, e.g., wetwood of Populus rre,nuloides ( Salica-
ceae) produces yellow fluoresceni streaks. Naturaily occurring fluorescence appears more uni -
form.
Oven-drying of samples or exposure to high temperatures or other extreme environmental
condirions may affect fluorescence propeilies.
WATER & ETHANOL EXTRACTS: FLUORESCENCE & COLOUR
205. Water extract fluorescent
206. Water extract basically colourless to brown or shades of brown
207. Water extract basically red or shades of red
208. Water extract basicalty yellow or shades of yellow
209. Water extract not as above
210. Ethanol extract tluorescent
211. Ethanol extract basically colourless to brown ar shades of brown
212. Ethanol extract basically red or shades of red
213. Ethanol extract basically yellow or shades of yellow
214. Ethanol extract not as above
Procedures:
Preparation of extracts: Add enough thin heartwood shavings to cover the bottom of a clean
vial which is approxirnately 20 mm x 70 mm. Do not use splinters or chips, because the extrac-
tion time is much longer than for shavings. For water extracts, cover the shavings to a depth of
approximately 20 mm (approximately 5 ml) with distilled water that is buffered at a pH of 6.86.
Packets of buffering agent are available from most scientific supply companies so that only the
contents of a packet need to be added to 500 or 1000 ml ofdistilled water to obtain the desired
pH. For ethanol extracts use 95% ethanol.
Dererminingfluorescence aí water & alcohol extracrs: Cover the vials and shake vigorous-
ly for 10 to 15 seconds. Allow the shavings and solution to stand for 1 to 2 minutes, and then
hold the vial under a longwave (approximately 365 nm) UV lamp and check for extract fluores-
cence. Generaily, extracts that fluoresce are bluish, but sornetimes they are greenish.
Determining colour of water and alcohol e.rrracts: Afrer determiriing the fluorescence of the
water and ethanol extracts, place the viais on a hotplate and bring the solution to a boi!. As soon
as the solution boils, remove the vial and immediately determine the colour.
Com,nenrs & Examples for fluorescence:
Examples of woods yielding water exrracts that fluoresce a hrilliant blue include Strychnos
decussata (Loganiaceae), Brosimum rubescens ( Moraceae), Olea europaea subsp. africana (Olea-
ceae), Prerocarpus indicar ( Papiionaceae), Zanrhaxylwnfiavwn (Rutaceae). Examples of wood
with weaker fluorescence of watercxtracts, but still positive, includeAcaciafarnesiana (Mimo-
saceae) and Lonc/iocarpus capassa (Papilionaceae).
Examples of woods yielding ethanol extracts with bright fluorescence include Protorhus lon-
gifolia (A naca rdi aceae), Cordia gerascanrhus ( Boragi naceae), A cacia eriolol'a (Mimosaceae).
IAWA Ust of microscopic features for hardwood identification 327
Examples of wood with weaker fluorescence of ethanol extracts, but still positive, include Kig-
gelaria africana (Flacourtiaceae), Acacia melanoxylon ( Mimosaceae), Olea capensis (Oleaceae).
Sometimes the water extract of a species fluoresces, bus its ethanol extract does not (e. g.,
Leucaena glauca - Mimosaceae). More often, the ethanol extract fluoresces, whi!e the water
extract does not (e. g., Afzelia quanzensis — Caesalpiniaceae, Lysiloma baha,nensis—Mimosa-
ceae).
Commenrs & Examples for water & alcohol exrract colour:
Water extract basically colourless tu brown or shades of brown is the most common of the
water extract feature colours.
Examples of woods with water extract basically red or shades of red (feature 206) include
Brasilettia spp. (Caesalpiniaceae), Catha edulis (Celastraceae), Cunonia capensis (Cunoniaceae),
and Mimusops caffra (Sapotaceae).
Examples of woods with water extract basically yellow or shades of ye!low (feature 2( y7) in-
clude Gonioma ka,na.ssi (Apocynaceae), Albina adianthifolia, Acacia caffra (Mimosaceae).
Feature 209, 'water extract not as above' includes colours such as orange, hlack, and purpie.
Various combinations of water extract colours occur as well.
Ethanol cxtract basically colourless to brown or shades ofbrown (feature 211) is the most
common.
Examples of woods with 'ethanol extract basicaily red or shades of red' (feature 212) include
Rhus integrifolia ( Anacardiaceae), l3aikiaea plurijuga, Peltophorum dubium, Swartzia madagas-
cariensis (Caesalpiniaceae), and Berchemia discolor (Rhamnaceae).
Exarnples of woods with 'ethanol extract basically yellow or shades ofyellow' (feature 213)
include Gonioma kamassi (Apocynaceac), Praeroxylon obliquwn, Zanrhoxylwnflavum ( Ruta-
ceae), Balanites ,naugharnii (Zygophyllaceae).
Feature 214, 'Ethanol extract not as above', includes co!ours such as orange, black, and
purp!e.
For more information, see Dyer(1988) and Quirk (1983).
FROTH TEST
215. Froth test positive
Procedure:
FolIow the procedure for preparing for lhe water extract tests. Add enough thin heartwood
shavings to cover the bottom of a clean via! approximately 20 mm x 70 mm. Do not use splinters
or chips, because the extraction time is much longer than for shavings; if sawdust is used, ex-
traction time will be less. For water extracts, cover the shavings to a depth of approximate!y
20 mm (approximately 5 ml) with distilled waler that is buffered at a p11 of 6.86. Packets of
huffering agent are available from most scientific supply companies so that only the contents of
the packet need to be added to 500 or 1000 ml of distilled water tu obtain the desired pH.
Cover the vial and shake vigorously for 10 to 15 seconds. If natural saponins are present in
large amounts, tiny bubb!es or 'froth' (like foam on a g!ass of beer) will be formed. A!low the
vial to stand for approximate]y 1 minute from the end of the shaking. If 'froth' still completely
covers the surface of the solution after 1 minute, the test is positive. If 'froth' or bubbles form
and then disappear within 1 minute, the test is negative. Ifonly some 'froth' remains around the
edge of lhe vial (i, e., forming a ring of 'froth'), but does not cover the entre surface, the test is
weakly positive.

328 IAWABuiletin n.s., Vol. 10 (3), 1989
Com,nen:s & Exa,'nples:
Positive 'froth' test reactions are produced by, e. g., Mora spp. (Caesalpiniaceae), Enrerolo-
bium cyclocarpum, Lysiloma baharnensis, Pseudosamanea spp. (Mimosaceae), Dipholis spp.,
Mastichodendron spp. (Sapotaceae).
Wcaldy positive reactions (ring of froth) are produced by, e. g., Pelrophorum spp. (Caesal-
piniaceae), Kiggelaria spp. (Flacourtiaceae), Ekebergia spp., Enrandrophragma spp. (Meia-
ceae), Acacia nigrescens (Mimosaceae), Millerria spp. (Papilionaceae), and llerchemia spp.
(Rhamnaceae).
For more information, see Dyer (1988), Quirk (1983), and Cassens and Milier (1981).
CITROME AZUROL-S TEST
216. Chrome Azurol-S test positive
Procedure:
Prepare a 0.5% solution of chrome azurol-S reagent by dissolving 0.5 g of the dry chrome
azurol-S granules and 5.0 g of sodium acetate (buffer) in 80 ml of distiiled water. After the
chemicais are compietely clissolved, add enough distilled water to make 100 ml of reagent. This
solution is stable and can be used over a number of years. To test dry wood samples, use ali
eyedropper to appiy one or two drops of the solution to a freshly exposed end-grain.
In highly positive woods, a bright blue colour wii develop in a matter of minutes, e.g., Poga
spp. (Anisophylleaceae), Cardwellia spp. (Proteaceae), Symplocos spp. (Symplocaceae), ali
Vochysiaceae. In those woods which absorb the solution very slowiy, e.g, Anisophyllea spp.
(Anisophylleaceae), Goupia spp. (Goupiaceae) or contain low concentrations of a]uminum, e.g.,
Laplacea spp. (Theaceae), Henrieriea spp. (Melastomataceae), several hours niay be required for
the blue colour to develop.
Comment:
Chrome Azurol-S tests for the presence of aluminum in both heartwood and sapwood. For
more information on this test, see Kukachka and Miller (1980).
Caution: Avoid decayed wood because chrome azurol-S is an indicator for some types of wood
decaying fungi.
BIJRNTNG SPLINTER TEST
217. Splinter burns to charcoal
218. Splinter burns to a fulI ash: Colour of ash bright white
219. Splinter burns to a fuIl ash: Colour of ash yellow.brown
220. Splinter burns to a fuil ash: Colour of ash other than above
221. Splinter burns to a partial ash
Definitions:
Charcoal = the blackened and charred remains of a splin ter, which usually burned slowly
andfor with difficulty, or the black and charred remnant of Lhe splinter with a fine thread of
black or grey ash which may remam attached.
IAWA List of microscopie features for hardwood identification 329
Fuil or complete ash = ash more or less retaining the shape of the original splinter.
Partial ash = ash that shrinks in size in comparison to the original splinter, lias a tendency
to drift away, and usually feels gritty when rubbed between the fingers.
Procedure:
Prepare match-size (approximately 2 x 2 x 50 rum) splinters from sound outer heartwood,
insure the wood is aI least air-dry. The splinter must be ignited with a match, and devices (e. g.,
Iighters) producing higher temperatures must be avoided. Ignite the splinter while it is held in a
vertical position wiffi a pair of tweezers/forceps. While the splinter is burning, hold it in a hoij
-zontalpsidurowly.
Some timber species will hurn with relative case (e. g., Populus spp. - Saiicaceae), while
others rnay show considerable reluctance (e. g., Eucalyptus paniculara - Myrtaceae). If it ap-
pears that the flame will extinguish before the splinter lias burned fully, combustion may be
aided by gently returning the splinter to a vertical position and then back to horizontal.
After Lhe flame extinguishes, it is important to aliow the giowing part of the splinter to extin-
guish before placing Lhe remnant on a cold surface.
Comments:
Certain splinters may crackle or produce bright sparks (e. g., Termina/ia carappa - Combre-
taceae), whilc others may produce a characteristic smoke coloration (heavy black smoke in Fim-
dersia iaevicarpa - Rutaceae) or exude coloured compounds while Lhey burn. Ali these features
rnay be recorded ia a description.
The descriptive classifications for appearance of the burnt splinter are those first recommend-
cd by Dadswell and Burneli (1932). In some iiterature, buff is used to describe spiinters that
have Lhe colour of pale tanned leather, a yellowy brown, e. g., Eucaiyptus panicuiara (Myrta-
ceae).
Apart from its use in CS1RO keys, Anonymus (1960) lias impiemented the feature and sug-
gests that it is of little vaiue except in distinguishing between some timbers which are closely
related anatornically.
For further information on the burning splintcr test, which so far has only been used on a
very limited scale, see Mann (1921), Welch (1922), Swain (1927), Dadswell and Burncli
(1932), and Mennega (1948).
Caution: Code greyish-white ash as 'other than above', as the white is reserved for obviously
(bright) white ash.
REFERENCES
Amos, G.L. 1952. Silica in timbers. C.S.T.R.O. Australia, Buii. No. 267.
Anonymus. 1960. identification of hardwoods. A lens key. For. Prod. Res. Buil. No. 25, Lon-
don.
Avelia, T., R. Dechamps & M. Bastin. 1989. Fluorescence study of 10,610 woody species
from Lhe Tervuren (Tw) coilection, Belgium. IAWA Buil. n. s. 9: 346-352.
Baas, P. 1973. The wood anatomical range in ilex and its phylogerietic and systematic signifi-
cance. Blumea 21: 193-258.
- 1986. Terminology of imperforate tracheary elements - in defence of libriform fibres with
minutely bordered pits. IAWA Buil. n. s. 7:82-86.

330 - -- lAWABulietin ms., Vol. 10(3), 1989
- , P.M. Esser, M.E.T. van der Westen & M. Zandee, 1988. Wood anatomy of Lhe Olea-
ceae. IAWA BulI. as. 9: 103-182.
- & M. Gregory. 1985. A survey of oil celis in Lhe dicotyledons with cornments on their re-
placement by and joint occurrence with mucilage celis. Israel J. of Bot. 34: 167-186.
- & F.H. Schweingruber. 1987, Ecological Irends ia lhe wood anatomy of crees, shrubs and
clirnbers from Europe. 1AWA Buli. n.s. 8:245-274.
Bailey, 1W, 1933. The canibium and its derivative tissues. VIII. Structure, distribution and
diagnostic significance of vestured pits in dicotyledons. J. Arnold Arbor, 14: 259-273.
- 1936. The problem of differentiation and c]assification of tracheids, 6ber-tracheids, and
libriform fibers. Trop. Woods 45: 18-23.
Braiier, J. & 01. Franklin, 1961. Identiflcatiort of hardwoods: a niicroscope key. For. Prod.
Res. Buil. 46. HMSO, London.
llridgwater, S. & P. Baas. 1982. Wood anatomy ofXanthophyllum Roxb. IAWA Buli. n.s. 3:
115-125.
Carlquist, S. 1980. Further concepts in ecological wood anatotny, with corrmients on recent
work in wood anatomy. Aliso 9: 499-553.
- 1985. Vasicentric tracheids as a drought survival mechanism in Lhe woody flora of southern
California and similar regions; review of vasicenlric tracheids. Aliso 11: 37-68.
- 1986a. Terniinology of imperforate tracheary elements. 1 AWA Buli, n.s. 7: 75-81.
- 1986b. Ibid.: A reply. IAWA Buil. as. 7:168-170,
- 1988. Comparative wood anatorny. Systematic, ecological, and evolutionary aspecis of
dicotyledon wood. Springer Verlag, Heidelberg, Berlin.
Cassens, D. L. & R. B. Milier. 1981. Wood anatomy ofthe New World Pithccellobium (sensu
lato). J. Arnold Arbor. 62: 1-44.
Chattaway, M.M. 1932. Proposed standards for numerical values used ia describing woods.
Trop. Woods, 29: 20-28.
- 1955. Crystals ia woody tissues I. Trop. Woods 102; 55-74.
- 1956. Crystals ia woody tissues II. Trop. Woods 104: 100-124.
Clarke, S.H. 1938, A multiple-entry perforated . card key with special reference to Lhe identifi-
cation of hardwoods. New Phytologist 37: 369-374.
ronquist, A. 1981. An inteated system of classification of fiowering plants. Columbia Univ.
Press, New York.
- 1988. The evolution and classification of fiowering plants. 2nd Ed, New York Botariical
Garden, Bronx, New York,
Dadsweu, 1 I. & M. Burneli. 1932. Method for lhe identification of coloured woods of the
genus Eucalyptus. CSIR Div. For. Prod., Technical Paper No. 5,
Dodd, R. S. 1986. Fiber Iength nieasurcrncnt systems: A review and modification of an existing
method. Wood & Fiber Sei. 18: 276-287,
Dyer, S.T. 1988. Wood fluorescence of indigenous South African trees. IAWA Buli, as. 9:
75-87.
Fahn, A., E. Werker & P. Baas. 1986. Wood anatomy and identífication aí trees and slirubs
from Israel and adjacent regions, Israel Acad. Sei. Hunianides, Jerusalem.
Fujii, T. 1988. Structure of latex and tanniniferous tubes in tropical hardwoods. BUI]. For.
Prod. Res. Inst. No. 352: 113-118 (+ 14 places; Japanese with English summary).
Gouwald, H. 1983. Wood anatomical studies of I3oraginaceae (s.1.). 1. Cordioidae. IAWA
Buli. n.s. 4: 161-178.
Gregory, M. & P. Baas. 1989- A survey of mucilage celis ia vegetative organs of the
Dicotyledons. Israel J. ofBot. 38: 125-174.
IAWA Ust of niicroscopic íeatiires for hardwood identification 331
Hart, C.A. & B. Swindcl. 1967. Notes on laboratory sarnpling of rnacerated wood fibers.
TAPPI 50 (7): 379 -3 8 1.
Hillis, H.E. 1987. Heartwood and tree exudates. Springer Verlag, Heidelberg, Berlin, London,
Paris.
IAWA Committee, 1964. Multi]ingual glossary of tenns used ia wood anatoniy. Konkordia,
Winterthur.
IAWA Comrnittee. 1981. Standard Iist ofcharacters suitable for coinputerized hardwood iden-
tilication. IAWA Bull. n. s. 2:99-110.
Ilic, Y. 1987. The CSIRO family card sorting key for hardwood identification. CSIRO Division
of Chemical and Wood Technology Technical Paper No. 8.
Kocppen, R. C. 1980. Silica bodies in wood of arboresccnt Leguminosae. IAWA Buil. n. s, 1:
180-184.
Kribs, D.A. 1968. Commercial forcign woods on Lhe American Market. Dover Publications,
New York.
Kukachka. B.F, & R.B. MilIer. 1980. A chemical spot-test for aluminum and its va.lue in wood
identification. IAWA Buli. n. s. 1: 104-109.
Mabberley, D.J. 1987. The plant-book, a portable dictionary of the higher plants. Cambridge
Univ. Press.
Mana, J. 1921. Ausiralian timbers, their strength, durability and identification. 2nd Ed., Wa]-
ker, May & Co. Melbourne.
Mennega, A.M.W. 1948. Suriname timbers 1. General Introduction, Guttiferae, Vochysiaceae,
Anacardiaceae and Icacinaceae. Martinus Nijhoff, The Hague.
Metcalfe, C. R. & L. Chalk. 1950. Anatomy oU the dicotyledons. Vol. I. Clarendon Press,
Oxford.
Mikesell, J. E. & R. H. Popham, 1976. Ontogeny and correlative relationship of the prirnary
thickening meristem in four-o'clock plants (Nyetaginaceae) maintained under long and short
photoperiods. Amer. J. Bot, 66: 997-1005.
Miller, R, B. 1981. Explanation of Lhe codingprocedure, IAWA Buli. as. 2: 111-145.
Qhtani, J., B. A. Meylan & B.G. Butterfleld. 1984. Vestures or warts - Proposed terniinology.
IAWA Buil. as. 5: 3-8.
Panshin, A.J. & C. DeZeeuw. 1980. Textbook of wood technology. 4th Ed.. McGraw-Hill,
New Yc,rk,
Parameswaran, N. & W. Liese. 1969. On Lhe formation and fine structure of septate wood
fibres of Ribes sanguineum. Wood Sei. Technol. 3: 272-286.
Parameswaran, N. & H.G. Richter. 1984. The ultrastructure of crystalliferous edis ia some
Lecythidaccae with a discussion of their terminology. IAWA Buil. n.s. 5: 229-236.
Quirk, J.T. 1983. Data for a coniputer-assisted wood identitication system. 1. Corarnercia]
legumes of tropical Asia and Australia. IAWA Buli. n. s. 4: 118-130.
Record, S.J. 1944. Keys to Anierican woods (continued). Trop. Woods 77: 18-38.
Richter, H.G. 1977. Differential staining of oil and mucilage ia idioblasts of Lauraccae. 1AWA
197714: 76.
- 1980. Occurrence, morphologv and taxonomic implications of crystalline and siliceous in-
cliisions in the secondary xylem of the Lauraceae and related farriilies, Wood Sci. Technol.
14: 35-44.
Sudo, S. 1959. Identification of Japanese woods. Buli. Govt. For. Exp. Stadon No. 118.
Tokyo, Japan.
Swain, E, H. 1927. Universal index to wood. Dep. of Public Lands. Queeasland For. Buli.
No. 7.
 332 IAWA Bulietin n.s., Vol. 10 (3), 1989

Takhtajan, A. 1980. Outline of the classificatiori of the fiowering plants (Magnoliophyta). Bot.
Rev. 46: 226-359.
- 1987. [Systema Magnoliophytorum.1 In Russian. Officina editoria 'Nauka', Leningrad.
Thorne, R.F. 1976, A phylogenetic classification of the Angiospermae. Evolutionary Biol. 9:
35-106.
Topper, S.M.C. & J. Koek-Noorrnan. 1980. The occurrence of axial latex tubes in the secon-
dary xylem of some species of Artocarpus J.R. & G. Forster (Moraceae). IAWA Bu11.
n.s. 1: 113-119.
Vidal Gomes, A., L. Lopes Teixeira, E. Gomes Schaitza & R.M. Hofmeister. 1989. Perfora-
tion plates in vesseis of Citharexylum myrianthum Cham. (Verbenaceae). IAWA Buil. ri. s.
10: 27-34.
Vliet, G.J.C.M. van. 1976a. Radial vesseis in rays. IAWA Buli. 1976/3-35-37.
- 1976b. Wood anatomy of Rhizophoraceae. In: P. Baas, J. Bolton, & D. M. Catling (cds.),
Wood sr.ructure in biological and technological rcsearch: 20-75. Leiden Bot. Series 3.
Leiden Univ. Press.
1978. The vestured pits of Cornbretaceae and allicd families. Acta Bot. Neerl. 27: 273-
285.
- 1981. Wood anatomy of the palaeotropical Melastoinaraceac. Blumea 27: 395-462.
- & P. Baas. 1984. Wood anatomy and classification of the Myrtales. Ann. Mo. Boi Gard.
71: 783-800.
Weber, W.A. 1982. Mnemonic three-letter acronyms for the farnilies of vascular planrs: a de-
vice forore effectivc herbarium curation. Taxon 31: 74-88.
Welch, M.B. 1922. A method of identification of some hardwoods. J. Roy. Soc. New South
Wales 56: 241-248.
WeUe, B.J.H. ter. 1976. Silica grains in woody plants of the neotropics, especially Surinam.
In: P. Baas,J. Bolton, & D.M. Catling (eds.), Wood structure in biological and techno-
logical research: 107-142. Leideri Bot, Series 3. Leiden Univ. Press.
- 1980. Cystoliths in the secondary xylem ofSparattanthelium (Hemandiaceae). IAWA Buil.
n.s. 1: 43-48.
Wheeler, E.A. 1986. Vesseis per square millimetre or vessel groups per square millimetre?
IAWA Buli. n.s. 7: 73-74.
- , R. G. Pearsori, C.A. LaPasha, T. Zack, & W. Hatley, 1986. Computer-aided wood iden-
tification. Reference manual. N.C. Ag. Res. Serv. Buli. 474.
Willis, J.C. 1973. A dictionary of the fiowering plants and ferns. 8th Ed., revised by H. K.
Airy Shaw. Cambridge Univ. Press.
 IAWA LIST OF MICROSCOPIC FEATUIRES FOR
HARDWOOD IDENTIFICATION

by a Cornmittee of he
International Association 01 Wood Anatorniqs

Veronica Angyalossy Alfonso, São Paulo, Brazil

Pieter Baas, Leiden, The Netherlands
Sherwin Carlquis, Claremont, California, USA
Joao Peres Chimelo, São Paulo, Brazil
Vera T. Rauber Coradin, Brasilia, Brazil
Pierre Dtienne, Nognt-surMarne, France
Peter E. Gasson, Km, UK
Dietger Grosser, München, FRG
Jugo flue, Highett, Victoria, Australia
Kc co Kuroda, Kyoto, Japan
Regis B. MiLer, Madison, Wiscons3n, USA
n Ogata, Tsukuba, Japan
Hans 'ieorg Richte:, Hamburg, FRO
I3en J. H. zer Wlle, Utrecht, The Netherlands
E1isabefl A. Wheeler, Raleigh, North Carolina, USA

Contenis

Preface ................................... 221

cknow1edgernents ............................ 223
Explanatory Notes............................ 225
..............................
List of Features 226
Name .................................... 233
natomical Features........................... 234
.............
Apper;dix —Non-anatornl Inforrnation 321
eferences ................................ 329

c ei; Sweeiiafruiicoa (Ferreir(.a spectabilis, Papilionaceae), n-ansver.se sec

on, x 115. Photograph by couriesy of eter E. Gasson (Kew, IJK) and taken
from a slide by Vera T. Rauber Coradin (Brasilia, Brazil).