University of Richmond UR Scholarship Repository

Biology Faculty Publications Biology

10-1984 Wood Anatomy and Relationships of Neowawraea (Euphorbiaceae) W. John Hayden University of Richmond, [email protected]

Dorthe S. Brandt

Follow this and additional works at: http://scholarship.richmond.edu/biology-faculty-publications Part of the Botany Commons, Horticulture Commons, Other Plant Sciences Commons, and the Plant Biology Commons

Recommended Citation Hayden, W. John, and Dorthe S. Brandt. "Wood Anatomy and Relationships of Neowawraea (Euphorbiaceae)." Systematic Botany 9, no. 4 (October 1984): 458-66.

This Article is brought to you for free and open access by the Biology at UR Scholarship Repository. It has been accepted for inclusion in Biology Faculty Publications by an authorized administrator of UR Scholarship Repository. For more information, please contact [email protected]. Systematic Botany (1984), 9(4): pp. 458-466 © Copyright 1984 by the American Society of Plant Taxonomists Wood Anatomy and Relationships of Neowawraea (Euphorbiaceae)

w. JOHN HAYDEN and DORTHE s. BRANDT Department of Biology, University of Richmond, Richmond, Virginia 23173

ABSTRACT. Wood anatomy of three specimens of Neowawraea phyllanthoides Rock, a rare and endangered member of Euphorbiaceae endemic to the Hawaiian Islands, is described and compared with woods of other genera of subfamily Phyllanthoideae. Neowawraea has often been associated or synonymized with Drypetes Vahl. Wood of Neowawraea is diffuse porous, perforation plates are simple, imperforate tracheary elements are thin-walled septate fiber-tracheids, rays are heterocel­ lular and crystalliferous, and axial xylem parenchyma is restricted to a few scanty paratracheal and terminal cells. In several respects these results differ from earlier published descriptions of the wood of this taxon; these earlier descriptions are shown to be in error. Further, as described here, wood of Neowawraea differs greatly from that of Drypetes. Accordingly, generic status for Neowaw­ raea is warranted on anatomical grounds and relationship with certain genera of tribe Phyllantheae such as Breynia Forst., Glochidion Forst., or Phyllanthus L. is favored over relationship with Drypetes.

Neowawraea Rock is a monotypic genus of ment tacitly implies a close relationship be­ dioecious trees endemic to the Hawaiian Is­ tween Neowawraea and Drypetes. lands. Historically recorded from Kauai, Oahu, In the course of surveying wood anatomy of Molokai, Maui, and Hawaii (St. John 1973), the Drypetes for comparison with several genera of species is currently considered in danger of ex­ subfamily Oldfieldioideae (Hayden 1980), it tinction (Fosberg and Herbst 1975). Given its became apparent that certain specimens re­ scarcity, and the fact that female flowers have ceived as "Drypetes phyllanthoides (Rock) Sherff," never been described (St. John 1973), it is not i.e., Neowawraea, were anatomically discordant surprising that proper understanding of its re­ with other woods of Drypetes. Furthermore, lationships has yet to be achieved. structures present in the three specimens avail­ Neowawraea phyllanthoides was discovered by able for study were also discordant with an ear­ Joseph Rock (1913) on rough aa lava flows of lier published description of the wood of "Dry­ Mauna Loa. Rock described this new genus and petes phyllanthoides" (Smith and Ayensu 1964). species on the basis of male flowers and mature Accordingly, the following account of wood of fruits, the only material known at that time. Neowawraea phyllanthoides is presented in an ef­ Concerning relationships, Rock stated, "it is fort to clarify the anatomy and relationships of evidently related to Phyllanthus"; he also per­ this rare tree. ceived a similarity to the baccate fruits of Bis­ chofia Blume. Pax and Hoffmann (1931) de­ MATERIALS AND METHODS clined to commit an opinion as to relationships, Specimens examined are listed in the Appen­ merely placing Neowawraea in a list of doubtful dix; xylarium acronyms follow Stern (1978). or incompletely described genera. Sherff (1939) Standard techniques were employed in prepa­ transferred this species to Drypetes Yahl, with­ ration of woods for microscopic examination. out citing any evidence for this change in sta­ Specimens with wide pores were embedded in tus; nevertheless this assignment has generally celloidin (Wetmore 1932) prior to sectioning, been followed by subsequent students of the others were sectioned without pretreatment Hawaiian flora (Neal 1965; St. John 1973; Fos­ beyond simple rehydration. Sections were pre­ berg and Herbst 1975; Carlquist 1980). Webster pared at thicknesses of 15, 20, and 25 µm and (1975), in a synopsis of a new classification of were stained with a combination of hematox­ Euphorbiaceae, maintained Rock's generic sta­ ylin (either Delafield's or Heidenhain's) and tus, but classified Neowawraea with Drypetes and safranin. Macerations were prepared according Putranjiva Wall. in his tribe Drypeteae (Griseb.) to Jeffrey's technique (Johansen 1940). Mea­ Hurusawa. Since Putranjiva is often included surements of anatomical structures were made within Drypetes (Hurusawa 1954; Airy Shaw in accordance with the procedures recom­ 1972, 1975, 1980; Willis 1973), Webster's treat- mended by Tamolang et al. (1963). Numerical 458 1984] HAYDEN & BRANDT: NEOWAWRAEA 459

TABLE 1. Numerical data for some features of wood of Neowawraea.

Mean number Mean pore Mean vessel Mean fiber- of pores per mm2 diameter (µm) element length (µm) tracheid length (µm) Specimen (N ~ 10) (N~SO) (N~ SO) (N~SO) Degener et al. 4371 54 53 541 857 St. John et al. 11354 116 35 605 795 Webster et al. 13889 58 50 545 871 values in the description of Neowawraea wood Rays heterocellular, very numerous, 14 per are averages of values from the three speci­ mm, 22 percent uniseriate, 53 percent multi­ mens studied; values included in parentheses seriate, 25 percent polymerous. Uniseriate rays are ranges. The variation from specimen to 18 (13-27) µm wide, 8.5 (3-23) cells or 520 (214- specimen of some features pertinent to adap­ 1809) µm tall. Multiseriate rays and multise­ tive morphology of xylem is noteworthy; data riate portions of polymerous rays 2.8 (2-4) cells for these features are presented separately as or 61 (37-87) µm wide; multiseriate ray height table 1. Otherwise, pooling of data in the wood 23 (9-36) cells or 955 (389-2714) µm; polymer­ description seems justifiable given: 1) the ous ray height 42 (19-118) cells or 1803 (442- small size of standard wood samples relative to 3919) µm. Prismatic crystals and tannin-like de­ the total amount of wood in a tree and, espe­ posits frequent. Vessel to ray pits circular and cially, relative to the dimensional variability of alternate to scalariform. wood elements from different positions in one Axial xylem parenchyma not abundant; dis­ tree (Panshin and De Zeeuw 1970); 2) the tribution scanty paratracheal and widely spaced small number of specimens available; and 3) terminal apotracheal. Parenchyma strand length the relative homogeneity of measurements from 7.6 (4-10) cells or 510 (369-663) µm. Vessel to the specimens studied other than those fea­ parenchyma pits circular and alternate to widely tures listed in table 1. spaced.

WOOD ANATOMY OF DISCUSSION NEOWAWRAEA PHYLLANTHOIDES ROCK Comparison of published data on the wood Figures 1-8. Growth rings present, faint, de­ structure of Neowawraea (as Drypetes phyllan­ fined by scattered terminal axial xylem paren­ thoides) with the results reported above reveals chyma and tangential alignment of first-formed several significant differences. Smith and Ay­ early wood pores. Porosity diffuse; pore groups ensu (1964) reported the absence of growth 25 percent solitary, 52 percent radial multiples, rings, a pore distribution of 95 percent solitary 22 percent clusters; 3.4 (2-12) pores per multi­ and 5 percent radial multiples, scalariform per­ ple pore group; 76 (54-116) pores per mm2 (see foration plates with 8 (6-11) bars per perfora­ table 1); pore area 0.114 mm2 per mm2 cross­ tion, libriform wood fibers (with no mention sectional area of wood. Pore outlines circular; of septa), and axial xylem parenchyma distri­ wall thickness 3 (1-5) µm; tangential diameter bution as diffuse and forming narrow bands; 46 (14-95) µm (see table 1). Perforation plates none of these features could be confirmed in simple; end walls inclined 51 (15-75) degrees our material despite the fact that two of our from the horizontal. Intervascular pits circular, specimens, USw 18672 and USw 18675, were also alternate; diameter 4.5 (2-6.6) µm. Vessel ele­ studied by Smith and Ayensu. On the other ments 564 (348-797) µm long (see table 1), often hand, Smith and Ayensu's data on pore out­ ligulate, containing variable quantities of tan­ line, pore diameter, vessel element length, ves­ nin-like deposits, with tyloses present in De­ sel wall thickness, intervascular pitting, imper­ gener et al. 4371. forate tracheary element length, and ray Fiber-tracheids septate, usually with 3-4 sep­ features conform with our descriptions. All ta per cell, 841 (429-1420) µm long (see table three specimens studied by us (see Appendix) 1); lumen diameter greater than or equal to wall were similar to each other and offered no clue thickness; vertical diameter of pit borders 2.4 as to the source of results so different from the (1-4.5) µm. earlier literature. Microscope slides of wood of 460 SYSTEMATIC BOTANY [Volume 9

Fies. 1-4. Wood structure of Neowawraea phyllanthoides (Webster et al. 13889). 1. Radial multiple and pore cluster, cross-section. 2. Heterocellular ray, radial section. 3. Septate fibers from radial section. 4. Uni­ seriate and biseriate rays, tangential section. All scales = 50 µm.

Drypetes at USw were examined in an effort to N eowawraea, and differ in the same respects resolve these discrepancies. As expected, ma­ from Smith and Ayensu's description. How­ terials labeled "USw 18672" and "18675" were ever, another slide labelled "Drypetes phyllan­ present; these conform with our description of thoides USw 18315" was also in the collection 1984] HAYDEN & BRANDT: NEOWAWRAEA 461

FIGS. 5-8. Wood structure of Neowawraea phyllanthoides (Webster et al. 13889). 5. Vessel element from macerated wood. 6. Vessel to ray pitting, radial section. 7. Crystal casts in macerated ray cells. 8. Alternate intervascular pits and multiseriate ray, tangential section. All scales = 50 µm. and the sections on it contain structures that men was accessioned from MADw, USw 18315 could account for the disputed descriptions is actually the African Drypetes gossweilleri S. from Smith and Ayensu. However, according Moore. Thus, it seems that parts of Smith and to data placed on file at USw when this speci- Ayensu's (1964) wood description of Neowaw- 462 SYSTEMATIC BOTANY [Volume 9

TABLE 2. Comparison of wood features of Neowawraea with other woods of Euphorbiaceae subfam. Phyl­ lanthoideae.

Character Aporusa-type Neowawraea Glochidion-type Perforations scalariform, or simple simple mixed scalariform and simple Im perforate non-septate septate septate tracheary elements thick-walled thin-walled thin-walled Axial xylem abundant not abundant not abundant parenchyma diffuse, diffuse- scanty paratracheal, scanty paratracheal in-aggregates, some terminal short bands Genera investigated Drypetes Neowawraea Bischofia Putranjiva Breynia Glochidion raea are erroneous as a result of a mislabelled The fact that N eowawraea possesses Glochidi­ preparation. Smith and Ayensu's documenta­ on-type structure suggests certain taxonomic tion of USw 18672 is also incorrect; the collec­ conclusions. Drypetes possesses Aporusa-type tion number is St. John et al. 11354 (US!), not wood, very different from that of N eowawraea; 11345. It does not seem likely, however, that thus Sherff's (1939) union of these genera is this second error has any bearing on the erro­ inappropriate and Rock's generic concept neou1tdescriptions. Our correction of the errors should be maintained. Further, inclusion of regarding perforation plates, imperforate tra­ Neowawraea in Webster's (1975) tribe Drype­ cheary elements, and axial xylem parenchyma teae (consisting of Drypetes, Putranjiva, and of N eowawraea is especially significant since Neowawraea) seems unwarranted. Both Drypetes these features hav~ considerable importance in and Putranjiva have similar wood structure, as the systematic anatomy of Euphorbiaceae recorded in Metcalfe and Chalk (1950) and con­ subfamily Phyllanthoideae. firmed here; that these entities are very closely For the most part, woods of Phyllanthoideae related is shown by the frequent inclusion of can be grouped into two distinct categories or Putranjiva in Drypetes (Hurusawa 1954; Airy structural syndromes. This was first recognized Shaw 1972, 1975, 1980). As a member of Dry­ by Janssonius (1934, 1950), who placed the peteae, Neowawraea is anatomically out of place. phyllanthoid genera he studied into the first The fossil record for phyllanthoid woods of­ and third of the four groups he recognized for fers an interesting perspective on relationships Javan euphorbiaceous woods. Metcalfe and among woody Phyllanthoideae. Several fossil Chalk (1950) added several genera to these cat­ woods with Glochidion-type structure have been egories, and referred to them as "Glochidion-type described from Upper Cretaceous deposits; woods" and "Aporosa (sic)-type woods." As can these include Paraphyllanthoxylon arizonense be seen from table 2 and figures 9-12, these from the southwestern United States (Bailey two structural syndromes are easily distin­ 1924) and Paraphyllanthoxylon capense, Securine­ guished on the basis of perforation plates, im­ goxylon biseriatum, and Bridelioxylon fibrosum from perforate tracheary element characteristics (see South Africa (Madel 1962). Assuming their as­ also Bamber 1974), and amount and distribu­ signment to Euphorbiaceae is correct, these fos­ tion of axial xylem parenchyma. N eowawraea sils indicate that Glochidion-type woods have clearly possesses all the characteristics of Glo­ been in existence for at least 60 million years. chidion-type woods and lacks the defining fea­ The great antiquity of Glochidion-type wood tures of the Aporusa-type. The significance of structure in Euphorbiaceae suggests that extant our redescription of the wood of Neowawraea genera bearing this structural syndrome should is thus apparent; based on Smith and Ayensu's not be classified with genera of the Aporusa­ (1964) description, Neowawraea would have fit type. easily in the Aporusq-type category. A search of Phyllanthoideae for plants more 1984] HAYDEN & BRANDT: NEOWAWRAEA 463

FIGS. 9-12. Wood structure of Drypetes (Aporusa-type) and Glochidion (Glochidion-type). 9. Drypetes ama­ zonica (Krukoff 6176), thick-walled fibers and abundant diffuse axial xylem parenchyma, cross section. 10. Drypetes lateriflora (Stern & Brizicky 558), scalariform perforation plates, radial section. 11. Glochidion borneense (Si Boeeca 1852), thin-walled fibers, cross section. 12. Glochidion borneense (Si Boeeca 1852), rays and septate fibers, radial section. All scales = 50 µm. closely related to Neowawraea than Drypetes ruseae (Lindl. ex Miq.) Airy Shaw, and Dry­ yields several possible genera, but no single peteae, can be eliminated on the basis of Apo­ definitive closest relative. Webster (1975) rec­ rusa-type wood structure in their constituent ognizes thirteen tribes within subfamily Phyl­ genera according to data contained in Metcalfe lanthoideae. Five of the thirteen tribes, Wielan­ and Chalk (1950). Five of the remaining tribes dieae Baill. ex Hurusawa, Amanoeae (Pax &: can be eliminated by their possession of var­ Hoffm.) Webster, Dicoelieae Hurusawa, Apo- ious unique morphological features (mostly 464 SYSTEMATIC BOTANY [Volume 9 presumed apomorphies) not shared with Neo­ ic status for Neowawraea within Phyllantheae wawraea: Bridelieae Muell. Arg., valvate sepals; should present no problem. Poranthereae (Muell. Arg.) Grilning, porous Of all the tribes of Phyllanthoideae, Phyllan­ dehiscence of anthers and herbaceous to eri­ theae thus seems able to accommodate N eo­ coid sub-shrub habit; Spondiantheae Webster, wawraea with the least amount of discord;. wood biseriate perianth and capsular fruit (probably structure and known reproductive features are plesiomorphies); Hymenocardieae (Muell. Arg.) broadly consistent with such a placement. Clas­ Hutch., samaroid fruits; and Uapaceae (Muell. sification in Phyllantheae is also consistent with Arg.) Hutch., globose involucrate inflores­ Rock's (1913) original opinion about the rela­ cences (morphological data from Pax and Hoff­ tionships of his genus. However, as noted man 1931). Three tribes remain: Antidesmeae above, Phyllantheae is anatomically heteroge­ (Endl.) Hurusawa, with six genera; Phyllan­ neous, so its generic composition may be con­ theae Dumort, consisting of 18 genera; and the sidered provisional at best. Ignorance of female monotypic Bischofieae Muell. Arg. Bischofia floral structures of N eowawraea further limits Blume possesses Glochidion-type wood struc­ the certainty of our conclusions. We therefore ture, but the genera of Antidesmeae and Phyl­ tentatively conclude that Neowawraea should be lantheae are heterogeneous-both Aporusa-type placed in tribe Phyllantheae, at least until fe­ and Glochidion-type woods are found in each male flowers of Neowawraea and further com­ (data from Metcalfe and Chalk 1950). Perhaps parative data for Phyllanthoideae become greater refinement is needed in the classifica­ available. tion of these anatomically heterogeneous tribes. ACKNOWLEDGMENTS. The research upon which this As possible relatives of Neowawraea, these three publication is based was supported in part by grants remaining tribes yield Bischofia, Antidesma L., from the University of Richmond Faculty Research Breynia Forst., Glochidion Forst., and Phyllanthus and Undergraduate Research Committees. Thanks are L., all of which possess Glochidion-type struc­ offered to curators of Aw, MADw, MAD-SJRw, and ture and are widely distributed on Pacific is­ USw for providing wood specimens. lands. Based on present evidence, wood anat­ omy offers no means of resolving a candidate LITERATURE CITED for closest relative to N eowawraea from these AIRY SHAW, H.K. 1965. Diagnoses of new families, five genera. new names, etc., for the seventh edition of Wil­ Female flowers and other comparative data lis's "Dictionary." Kew Bull. 18:249-273. for Neowawraea are sorely needed, since the data ---. 1967. Notes on the genus Bischofia Bl. (Bis­ currently available allows only partial resolu­ chofiaceae). Kew Bull. 21:327-329. tion of its relationships. However, present --. 1972. The Euphorbiaceae of Siam. Kew Bull. knowledge does allow some speculation con­ 26:191-363. cerning relationships with the five genera --. 1975. The Euphorbiaceae of Borneo. Kew Bull. Additional Series 4:1-245. identified above. Antidesma and Bischofia are ex­ --. 1980. The Euphorbiaceae of New Guinea. cluded routinely from Euphorbiaceae by Airy Kew Bull. Additional Series 8:1-243. Shaw (1965, 1967, 1975, 1980; see also Willis BAILEY, I. W. 1924. The problem of identifying the 1973), the former on the basis of its unilocular wood of Cretaceous and later dicotyledons: Para­ drupaceous fruits with distinctive foveolate-re­ phyllanthoxylon arizonense. Ann. Bot. (London) 38: ticulate and often flattened endocarps, and the 439-451. latter because of its trifoliolate to pinnately BAMBER, R. K. 1974. Fibre types in wood of Eu­ compound leaves. Neither of these problemat­ phorbiaceae. Austral. J. Bot. 22:629-634. ical genera seems especially close to Neowaw­ CARLQUIST, S. 1980. Hawaii a natural history. Hono­ raea. Only Breynia, Glochidion, and Phyllanthus lulu: Pacific Tropical Botanical Garden, SB Print­ remain as reasonable relatives of Neowawraea. ers, Inc. FOSBERG, F. R. and D. HERBST. 1975. Rare and en­ These three genera are all closely related mem­ dangered species of Hawaiian vascular plants. bers of Webster's tribe Phyllantheae. Breynia and Allertonia 1:1-72. Glochidion differ from Neowawraea by their cap­ HAYDEN, W. J. 1980. Systematic anatomy of Old­ sular fruits, and Phyllanthus differs in its fieldioideae (Euphorbiaceae). Ph.D. dissertation uniquely specialized growth habit; thus gener- (unpubl.), University of Maryland, College Park. 1984] HAYDEN & BRANDT: NEOWAWRAEA 465

HURUSAWA, I. 1954. Eine nochmalige Durchsicht des as available: locality, collector, collection number, herkommlichen Systems der Euphorbiaceen in xylarium accession number, and location of herbar­ weiteren Sinne. J. Fae. Sci. Univ. Tokyo, Sect. 3, ium voucher. Bot. 6:209-342. JANSSONIUS, H. H. 1934. Euphorbiaceae. Mikrogra­ phie des Holzes der auf Java vorkommenden Bischofia javanica Blume. New Guinea, USw 24030 (L); Baumarten 5:442-812. West Irian, USw 35137 (L); Fiji, Smith 6045, Aw 28452 --. 1950. Wood anatomy and relationship. Blu­ (A); Sumatra, Si Boeeca 3558, USw 28298 (MICH); mea 6:407-461. New Caledonia, USw 4622; New Caledonia, USw JOHANSEN, D. A. 1940. Plant microtechnique. New 4626. York: McGraw-Hill Book Company. Breynia cernua Muell. Arg. Sumatra, Si Boeeca 8029, MADEL, E. 1962. Die fossilen Euphorbiaceen-Holzer USw 29110 (MICH). mit besonderer Beriicksichtigung neuer Funde Breynia disticha J. R. Forst. & G. Forst. Brookside Bo­ aus der Oberkreide Siid-Afrikas. Senckenberg. tanical Garden, Wheaton, Md. (cultivated), Hayden Lethaea 43:283-321. 3 (MARY). METCALFE, c. R. and L. CHALK. 1950. Anatomy of the Breynia patens Benth. Sumatra, Si Boeeca 1413, USw dicotyledons. Oxford: Clarendon Press. 27761 (MICH). NEAL, M. C. 1965. In gardens of Hawaii. Honolulu: Drypetes afzelii Hutch. Liberia, Cooper 315, USw 4895; Bishop Museum Press. Ivory Coast, Detienne 135, MADw 36858 (P). PANSHIN, A. J. and c. DEZEEUW. 1970. Textbook of Drypetes alba Poit. Cuba, Gill & Whitford 105, USw wood technology, vol. l, 3rd ed. New York: Mc­ 2985 (NY). Graw-Hill Book Company. Drypetes amazonica Steyermark. Brazil, Krukoff 6176, PAX, F. and K. HOFFMANN. 1931. Euphorbiaceae. Pp. USw 7532 (NY). 11-235 in Die naturlichen Pflanzenfamilien, 19c, 2nd Drypetes cf. assamica Pax & Hoffm. Sumatra, Krukoff ed., A. Engler and K. Prantl. Leipzig: Wilhelm 4075, USw 6310 (NY). Engelmann. Drypetes australasica (Muell. Arg.) Pax & Hoffm. New ROCK, J. F. 1913. The indigenous trees of the Hawai­ Guinea, Pullen 6845, MADw 29077 (K). ian Islands. Honolulu: privately published. Drypetes aylmeri Hutch. & Dalz. Ivory Coast, Bamps SHERFF, E. E. 1939. Additional studies of the Hawai­ 2390, MADw 32770 (BR). ian Euphorbiaceae. Field. Mus. Bot. ser. 17:547- Drypetes bordenii (Merr.) Pax & Hoffm. Philippines, 576. Stern 2155, USw 31814 (US). SMITH, A. c. and E. s. AYENSU .. 1964. The identity Drypetes carolinensis Kanehira. Yap, Kanehira 1940, of the genus Calyptosepalum S. Moore. Brittonia MAD-SJRw 20419. 16:220-227. Drypetes diversifolia Krug & Urban. Florida Keys: Stern ST. JOHN, H. 1973. List and summary of the flow­ 2684 (MARY); Stern et al. 2765 (MARY); Stern et al. ering plants in the Hawaiian Islands. Pac. Trop. 2885 (MARY); Stern et al. 3369 (MARY). Bot. Gard. Mem. 1:1-519. Drypetes dolichocarpa Kanehira. Saipan, Kanehira 2141 STERN, W. L. 1978. Index xylariorum. Institutional (wood no. A2216), MAD-SJRw 26843. wood collections of the world. 2. Taxon 27:233- Drypetes glauca Vahl. Dominica, Stern & Wasshausen 369. 2524, USw 35555 (US). TAMOLANG, F. M., R.R. VALBUENA, J. A. MENIADO, and Drypetes lateriflora (Sw.) Krug & Urban. Florida Keys: B. C. DE VELA. 1963. Standards and procedures Stern & Chambers 266, MAD-SJRw 51473 (MAD); Stern for descriptions of dicotyledonous woods. La­ & Brizicky 558, MAD-SJRw 51330 (MAD); Stern 2683 guna, Philippines: Forest Products Research In­ (MARY). stitute Bull. 2. College. Drypetes ovalis Pax & Hoffm. Java, For. Dept. Java 2703, WEBSTER, G. L. 1975. Conspectus of a new classifi­ MAD-SJRw 22255 (L). cation of the Euphorbiaceae. Taxon 24:593-601. Drypetes ovata Hutch. Liberia, Cooper L4, USw 5695 WETMORE, R.H. 1932. The use of celloidin in botan­ (K). ical technic. Stain Technol. 7 :37 -62. Drypetes principum (Pax) Hutch. Gold Coast, Vigne WILLIS, J. C. 1973. A dictionary of the flowering 3423, MAD-SJRw 29718 (FHO). plants and ferns, 7th ed., Rev. H.K. Airy Shaw. Drypetes sclerophylla Mildbr. Tanganyika, Schlieben Cambridge: University Press. 5473 (wood no. 441), MAD-SJRw 30010 (MAD). Drypetes sumatrana Pax & Hoffm. Burma, USw 11226. APPENDIX. WOOD SPECIMENS EXAMINED Glochidion album (Blco.) Boerl. Philippines, Stern 2140, USw 31799 (US). Specimens examined in this study are documented Glochidion amentuligerum (Muell. Arg.) Croizat. Fiji, in the following list. Entries for each species include, Smith 1926, MAD-SJRw 28417 (MAD). 466 SYSTEMATIC BOTANY [Volume 9

Glochidion arborescens Blume. Sumatra, Si Boeeca 7714, Glochidion philippicum (Cav.) C. B. Robins. Philippines, USw 29032 (MICH). Stern 2150, USw 31809 (US). Glochidion borneense Boerl. Sumatra, Si Boeeca 1852, Neowawraea phyllanthoides Rock. Hawaii, Oahu, De­ USw 27909 (MICH). gener et al. 4371, USw 18675 (BISH); Hawaii, South Glochidion bracteatum Gillespie. Fiji, Gillespie 3530, Kona, Kapua, St. John et al. 11354, USw 18672·(BISH); MAD-SJRw 25675 (BISH). Hawaii, Maui, SW slopes of Haleakala, Webster et Glochidion cuspidatum var. samoanum (Muell. Arg.) al. 13889 (DAV). Pax. Samoa, Christophersen 3301, MAD-SJRw 26135 Putranjiva roxburghii Wall. West lrian, USw 34683 (L); (BISH). Trinidad (cultivated), Broadway 9248, AW 22185 (A).