The Platycarya Perplex and the Evolution of the Juglandaceae Scott L. Wing; Leo J. Hickey American Journal of Botany, Vol. 71, No. 3. (Mar., 1984), pp. 388-411. STOR ? Stable URL: http://links.jstor.org/sici?sici=0002-9122%28198403%2971%3A3%3C388%3ATPPATE%3E2.0.CO%3B2-A American Journal of Botany is currently published by Botanical Society of America. Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://www.jstor.org/about/terms.html. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use. Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at http://www.jstor.org/journals/botsam.html. Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission. JSTOR is an independent not-for-profit organization dedicated to and preserving a digital archive of scholarly journals. For more information regarding JSTOR, please contact support@jstor.org. http ://www.j stor.org Fri Dec 15 12:53:47 2006 Amer. J. Bot. 71(3): 388-411. 1984. THE PLATYCARYA PERPLEX AND THE EVOLUTION OF THE JUGLANDACEAE1 SCOTT L. WING AND LEO J. HICKEY U.S. Geological Survey, E-501 National Museum of Natural History, Washington, DC. 20560, and Peabody Museum of Natural History, Yale University, New Haven, Connecticut 06520 ABSTRACT We report on the leaves, fruits, inflorescences, and pollen of two fossil species in the genus Platycarya. The association of these dispersed organs has been established by their repeated co-occurrence at a large number of localities, and for two of the organs (fruit and pistilla^ inflorescence, and pollen and staminate inflorescence) by apparent organic attachment of compression fossils. Each of the two species can be distinguished by characteristics of all the known megafossil organs. We also review the fossil record of dispersed platycaryoid fruits and inflorescences, recognizing three additional species of Platycarya and two of Hooleya. Two of the fossil Platycarya species are morphologically very different from the living Platycarya strobilacea Sieb. et Zucc, but they show the diagnostic features of the genus. Hooleya is a generalized member of the Platycaryeae that is probably close to the ancestry of Platycarya. The two Platycarya species known from multiple organs provide a remarkable example of mosaic evolution in which fertile and foliar structures have attained different levels of mor- phological specialization. The leaves, often considered the most plastic of plant organs, retain several features that are otherwise seen only in the Engelhardieae. These similarities in leaf architecture between the fossil Platycarya species and Engelhardieae are advanced features for the Juglandaceae, and thus indicate a sister-group relationship between the two lines. In contrast to the leaves, the fruits, inflorescences, and pollen of the fossil Platycarya species are almost as specialized as those of the extant P. strobilacea and bear little resemblance to the same structures in other genera of the family. The morphology, taphonomy, sedimentary setting, and geographic and stratigraphic distri- bution of three of the fossil platycaryoid species suggest that they were wind-dispersed, early successional plants that grew in thickets. This habit is retained by Platycarya strobilacea and is typical of many of the amentiferae (e.g. Myricaceae, Betulaceae). The /--selected life-history pattern of the Platycarya line may well have contributed to its low diversity through geologic time. THE RELATIONSHIP of Platycarya to the rest unique in the Juglandaceae. Neither the gen- of the Juglandaceae has long remained an in- eralized conditions nor the unique specializa- tractable problem in the systematics of the wal- tions of the genus are of much value in deci- nut family. Manning (1938, 1940, 1948) con- phering the relationships between Platycarya eluded that the inflorescences and flowers of and the other genera of the family. Platycarya strobilacea2 Siebold et Zuccarini re- For similar reasons the fossil record has also tain many features that are primitive for the shed little light on intrafamilial relationships, family as a whole. However, P. strobilacea also Platycarya-tike pollen first occurs in the fossil has advanced features, such as pseudocolpate record at or slightly below the Paleocene-Eocene pollen, condensed pistillate inflorescences and boundary (Frederiksen and Christopher, 1978), spirally thickened vessel elements, that are but these early Tertiary grains already bear the characteristic pseudocolpi of Platycarya, and ' Received for publication 28 March 1983; revision ac- differ from pollen of P. Strobilacea chiefly in cepted 23 August 1983. being larger and in having a more variable We thank Peter R. Crane, Steven R. Manchester, Amy shape and arrangement of pseudocolpi. Four R. McCune, Doug J. Nichols, Harold E. Robinson, Bruce published megafossil taxa are attributed to H.TifFney and Jack A. Wolfe for discussion and criticism. Platycarya. two species based on leaves, P. James P. Ferngno, Frederick Meyer, Howard E. Schorn >u*yi-<*yi*.i F ' and Jack A. Wolfe made specimens available. Charles D. miocemca Hu et Chaney (1940) and P. pseudo- Blome provided facilities for photography of specimens brauni (Hollick) Wolfe (1979); and two species and Loraine utter drafted the figures. The first author based on fertile material, P. richardsonii(Bow- thanks Natasha Atkins for careful editing and endless pa- ^rbank) Chandler (1961) and P. americana "un this paper we follow Manning (l 978) in takingPtoy- Hickey (1977). These four species were initially carya strobilacea to include P. longipes Wu. described as being very similar to Platycarya 388 March, 1984] WING AND HICKEY?PLATYCARYA PERPLEX 389 strobilacea, and so made little contribution to understanding the relationships of the platy- caryoid line. This study differs from previous ones in two ways. First, for two fossil species of Platycarya we have studied several organs, including leaves, fruits, inflorescences and pollen. Sec- ond, the species of Platycarya and Hooleya covered here differ from Platycarya strobilacea in ways that provide a key to understanding the relationship of Platycarya to the other Ju- glandaceae. TAXONOMIC CONCEPTS?i^oiH'/.geHtfra?Prior to the publication of the 1978 Code of Botan- ical Nomenclature, paleobotanists recognized two types of genera as "holding categories" for remains whose biological affinities were un- clear. The first of these, the form genus, consists of remains unassignable to a family; the second type, the organ genus, constituted remains that could be so assigned (Stafleu et al., 1972). In 1978 the organ generic category was eliminated (Stafleu et al., 1978), nevertheless the feeling has persisted among many paleobotanists that genera based on fossils are in some way arti- ficial because the material is often fragmentary and does not represent the whole plant. In con- trast, we do not believe that fossil genera are necessarily artificial; the "naturalness" of a ge- nus depends on characters, not simply on how much of the plant body is known. Failure to make this distinction may lead to the indis- criminate use of form genera, in which are lumped both fossil taxa that are poorly known and those that are well known but morpho- logically generalized. In a cladistic sense, each natural genus is defined by a unique character or set of char- acters that is consistently found in its member species. For convenience in discussion we will consider fossil taxa in two groups: those that possess characters defining a natural genus, and those that lack such characters. The appropriate "holding category" for a dispersed organ in the first group (i.e., one that has at least one generic autapomorphy, a unique character diagnostic of a natural genus) is a natural genus. The chief argument against plac- ing a dispersed organ in a natural genus is that it creates nomenclatural instability (the even- tual discovery of more complete fossil remains may yield characters that would require the species to be transferred to another genus). However, the problem of nomenclatural in- stability is not as severe as might be thought. Additional organs of a fossil species that are recovered after the initial description can give three kinds of information. One, the newly found organ(s) may be similar to those of other species in the genus. If so the original generic placement is borne out. Two, the newly dis- covered organ(s) may be more generalized (ple- siomorphic) than are the same organs in other species in the genus. In this case the original generic placement is still correct because the fossil species possesses the diagnostic auta- pomorphy(-ies) of the genus; the fossil species is simply more primitive than its congenerics. (For an example in which a large number of primitive features might influence generic as- signment, see discussion of stem groups be- low.) Three, the newly discovered organ(s) may show new autapomorphies. If these autapo- morphies are significant, they may be used as the basis of a new natural generic category. Only this last kind of new information is likely to require a generic transfer. Even if the dis- persed organ had originally been placed in a form genus, rather than in a natural genus, the discovery of new generic-level autapomor- phies would require a redefinition of the form genus, or the naming of a new natural genus. Thus form genera do not appear to significantly increase nomenclatural stability for fossil taxa that have autapomorphies that should allow them to be placed in a natural genus. The dis- advantage of using form genera is that they relegate the fossil material to a category that has no clear place in the hierarchy of botanical nomenclature, a category that does not convey the relationships of the fossil as precisely as does a natural genus. What about fossil material falling in the sec- ond group, those that lack generic autapo- morphies? Some fossils lack generic autapo- morphies because they are extremely generalized and thus indistinguishable from the hypothetical ancestor of one or more other taxa in the family. For example, the Paleocene ju- glandaceous fruit Casholdia (Crane and Man- chester, 1982) has two synapomorphies with Engelhardia and Oreomunnea, but it differs from them by being more plesiomorphic in four other characters. This fossil lacks any ge- neric-level autapomorphies, yet it is easily dis- tinguished from other juglandaceous genera by its unique combination of plesiomorphies and apomorphies. Such morphological ancestors are best placed in a monotypic stem group (as Crane and Manchester have implicitly indi- cated for Casholdia). This practice expresses phylogenetic history and relationships far bet- ter than the use of form genera and also serves to distinguish generalized taxa from those that are just poorly known. Other fossils may lack generic autapomor- phies because they are poorly preserved or be- 390 AMERICAN JOURNAL OF BOTANY [Vol. 71 cause the organs that bear the significant fea- tures have not been recovered. For such fossils, two options are available: the fossil can be left without a formal name, or it can be placed in a form genus. Form genera are most useful when the fossil species is a clearly recognizable entity that has some biostratigraphic or paleo- ecological importance, even though its botan- ical affinities are unclear. In our opinion, how- ever, the form-genus category should be used only when affinities are unclear because diag- nostic apomorphies are missing or unavailable. It should not be used simply because the species is based on dispersed organs, or in species where the diagnostic characters are present but in the plesiomorphic character state. Although form genera are meant to apply when affinities at the family level cannot be determined, we find this stipulation arbitrary. The essence of the form-genus category is that it allows a formal name to be given to a species that is incertde sedis with respect to some more inclusive level in the hierarchy. Whether this more inclusive level is a tribe, family, or a class should make no difference. In conclusion, we find that fossil genera need not be considered artificial, and that, generally, the use of natural genera is preferable to the use of form genera because it more precisely expresses phylogeny. The use of form genera should be restricted to fossils for which a nat- ural genus cannot be defined or determined because key characters are unavailable as a result of some imperfection of the fossil ma- terial. Even under these circumstances it may be better to leave the fossil without a formal name unless it has some biostratigraphic or paleoecological utility. Association of dispersed organs?The study of multiple fossil organs belonging to the same biological species may bring the needs of no- menclatural stability and biological accuracy into conflict. The practice of placing each kind of dispersed fossil organ in its own taxon (at the generic or specific level) preserves nomen- clatural stability should the organs later prove to have been spuriously associated and pro- vides separate names for morphological enti- ties that are not usually found connected. How- ever, the use of such organ taxa may hinder the recognition of true biological entities (species), and may thus impede evolutionary or systematic syntheses based on the fossil re- cord. Therefore, we assert that dispersed or- gans should be included under a single name when there is strong evidence that the dis- persed organs belong to a single biological en- tity. TABLE 1. Occurrence of the dispersed organs o/Platycar- ya americana and P. castaneopsis at 26 localities. Symbols: P, present; P?, doubtfully present; A, absent; ?, no sample processed for pollen; ip, localities at which fossils were poorly preserved; * localities for which only small samples were available. Detailed locality information is given in Appendix 1 Pistil- Stami- late Leaf- Localities lets Fruits rescence rescence Pollen PLATYCAR YA AMERICANA USGS9559V A A P? A - USNM14053*^ A A P? .A - USNM14048 P P P P P USGS9394i/< P P P? P - PLATYCARYA CASTANEOPSIS UCMP PA114 P P P A ? UCMP PA 104 P P P A - USGS8912 P P P A P USGS9397 P P P A - USGS9051* P P A A - USGS9538* P P P A - USGS9179* P P P A - USGS8894* P A A A - 134309* P A A A P BCT P P P A - CQ P A A A P FL P A P A P WhBgi/- P A P A P TAT79f P A P A P BGT2^ P A P A - MBR* P A P A - SL4 P A A A - 15ME^ P A A A - Bi/< P A A A - 15Mi/< P A A A - 114 P A A A - TV- P A A A - WhBgD/- P A A A - 72^ P A A A - What evidence is necessary? Petrified ma- terial showing anatomical connection between parts is convincing but extremely rare, espe- cially in Tertiary floras. Attachment of struc- tures in compression may provide good evi- dence, depending on the quality of preservation and the amount of organic matter on the bed- ding surface. Physical proximity of dispersed organs in the sediment is generally considered weak evidence that the organs were parts of the same plant. However, we argue here that under special circumstances the physical prox- imity of dispersed plant organs can provide strong evidence that they belonged to the same plant and that placement of such dispersed or- gans in the same genus and species can be ad- equately justified. The association of dispersed organs in Platy- carya americana and Platycarya castaneop- sis?Platycarya americana is here emended to March, 1984] WING AND HICKEY?PLATYCARYA PERPLEX 391 120? 115? 110 Idaho 105 %? ?-'>,-? '??19 J5-14AJ ^%> ? 15-18 i K A3 '>.':??" ?20- 21'T-y^ ^ ">%\ ? 23 Wyoming ^ *22/Af ".: "c* 'i ,24 ^ ^L A 15% of the palynoflora) in all of them. Male inflores- cences of P. castaneopsis were not recovered, probably because they were too fragile to be readily fossilized. Quarries that produced few- er than three megafossil organs were localities where preservation was relatively poor and/or collections were small (Table 1). Three arguments can be advanced to explain the repeated physical association of the same plant parts at different localities: 1) the plant material was sorted by wind or water (i.e., the dispersed organs are hydrodynamically simi- lar); 2) the plants of two or more species that produced the organs all grew in the same hab- itat; 3) the organs belonged to the same species of plant. For these fossil Platycarya species we can effectively eliminate hypotheses one and two. If the co-occurrence of the organs belonging to Platycarya is a result of water sorting pro- cesses, then we should expect to find evidence for current flow in the sediment. Field studies of localities where Platycarya occurs show that the fossils were all deposited in low-energy flu- vial or fluvio-paludal settings (Hickey, 1977; Wing, 1980, 1981, in press, and unpubl. field observ.). The fossiliferous sediment is silt or mixed clay and silt, and current features are generally absent. The plant megafossils are nei- ther torn nor folded. Like other low-energy fluvial settings (Scheihing and Pfeflerkom, 1982), there is extremely low potential here for 392 AMERICAN JOURNAL OF BOTANY [Vol. 71 800 700 600 500 400 300 200 100 PERCENT PLATYCARYA MACROFLORA PALYNOFLORA 40 30 20 10 30 20 10 Tattnan Formation ^ ?^ A- ?A \ \_ W AT IO ?- iZ D FO R - - . LW OC - 5 : t2 : ; 40 30 20 10 C ) 30 20 10 0 800 700 600 p-500 400 300 200 --100 Fig. 2. Stratigraphic occurrence of Platycarya micro- and megafossils in the Willwood Formation and Tatman Formation of the Bighorn Basin, Wyoming. Symbols: ?, sampled level where no Platycarya remains were found; ?, doubtful or uncertain occurrence of Platycarya; ?, pres- ent at the plotted percentage of the flora. the deposition and preservation of allochtho- nous plant remains. In addition, the dispersed organs of Platycarya (small, winged fruits; woody, cone-like pistillate inflorescences; leaf- lets; pollen) are not hydrodynamically similar to one another. The hypothesis that the dispersed organs of each Platycarya species actually belong to two or more species that are ecological associates is also controverted by the low energy of de- position and the high fidelity of preservation at the Platycarya localities. If the organs of each Platycarya complex were derived from more than one species, other plant megafossils should occur consistently with Platycarya (e.g., the leaves that really belong with the fruits of P. castaneopsis). In actuality, no kind of plant megafossil consistently occurs with organs of either Platycarya species (Appendix 2), and no other Juglandaceae have been identified at the Platycarya localities. As the two associations of dispersed organs that form P. americana and P. castaneopsis are unlikely to have been created by sorting or by the mixing of the organs of more than one species, the only remaining hypothesis is that each set of organs belonged to one plant. This hypothesis is further strengthened by the strati- graphic distribution of P. castaneopsis leaflets and pollen. The stratigraphic evidence is derived from work in the Willwood and Tatman Formations of the Bighorn Basin in northwestern Wyoming (Fig. 1). An 820-m-thick stratigraphic section was measured through these formations, and pollen samples were taken at frequent intervals (Wing, 1981). Pollen of Platycarya is sporadic and in low abundance through most of this section (Fig. 2). At 700 m it becomes extremely abundant and constitutes at least 15% of the palynoflora through the rest of the section. This pattern matches the stratigraphic distribution of P. castaneopsis leaflets, which have their lowest reliable occurrence at 700 m and dom- inate the megaflora from this level up (>25% in censused collections). The high abundance of platycaryoid pollen occurs at all six Platy- carya megafossil localities for which pollen data are available. Further evidence for considering the dis- persed organs as parts of the same species comes from the intrafamilial affinities of each organ type. Within each of the allopatric organ com- plexes representing P. americana and P. cas- taneopsis, the inflorescences, fruits, and pollen show features characteristic of the tribe Platy- caryeae, whereas the leaflets have features al- lying them with both Platycaryeae and Engel- hardieae. The consistent affinity of these organs to particular parts of one family further di- minishes the probability that they are derived from more than one kind of plant. The final line of evidence for the association of these organs is organic connection in some compression fossils. Material from the Wind River Formation (University of California Museum of Paleontology [UCMP] locality PA 114) shows the small, winged fruits of P. cas- taneopsis borne in cone-like inflorescences of the type usually found in close proximity to them (Fig. 16, 21, 22). In material from the Golden Valley Formation (U.S. National Mu- seum [USNM] locality 14048; Fig. 10, 12), anthers of the staminate inflorescences of P. americana contain the same type of platycar- yoid pollen that dominates the dispersed pollen floras of localities that produce Platycarya megafossils. Given the overwhelming evidence from re- peated physical association, sedimentary con- text, stratigraphic distribution, taxonomic af- finity, and physical connection, we feel fully justified in treating these two complexes of dis- persed fossil organs as biological entities, and in placing the organs of each complex under one name. Because of the strong evidence that each complex of organs represents one kind of March, 1984] WING AND HICKEY?PLATYCARYA PERPLEX 393 plant, the advantages to systematic and evo- lutionary discussion of using one name far out- weigh the risk of nomenclatural instability. SYSTEMATICS SECTION Genus Platycarya Siebold et Zuccarini 1843 Emended description: Description in italics refers to conditions observed in fossil species; leaf architecture terms after Hickey (1979). Leaves odd-pinnately compound with 5-11 leaflets; leaflets with short to long petiolules and rounded, semicordate, acute or decurrent bases that may have patches of hair or show some elaboration and enrollment of the basi- laminar tissue; margin serrate; teeth Cunonioid (Hickey and Wolfe, 1975), with sharp or round- edweakly glandular apices; secondaries leaving the midvein at 65? to 90? and curving gently or sharply upward near the margin; interse- condaries weakly to strongly developed; ter- tiary veins percurrent, straight or retroflexed, making perpendicular or obtuse junctions with the subadjacent secondary; quaternary vein or- der easily or not easily distinguished from higher order veins; areoles 0.15-0.30 mm in diame- ter, weakly or strongly impressed, and with 50/o-10?/o-90% of the areoles having freely end- ing veinlets; marginal venation with two weak or strong veins ascending from each sinus to the superadjacent tooth. Inflorescence axes usually bisexual (mature pistillate axes bearing lateral scars below the fertile part) with the staminate inflorescences lateral to the pistillate. Pistillate inflorescences 3-8 cm long and 1-2 cm wide, cone-like with woody bracts arranged spirally along the distal part of the axis; bracts 5-11 mm long and 2- 5 mm wide, with acute to rounded apices; per- sistent on the axis. Fruits 2-10 mm wide and 2-5 mm high (proximal-distal direction); bilaterally sym- metrical, nutlets winged; lateral wings formed of fused bracteoles and lateral sepals; median sepals reduced, or persistent and usually ex- tending past the stigmatic region; stylar arms carinal, short (<0.5 mm), stigmatic surface adaxial, or partly terminal; nutlet attachment oriented abaxially, basal/abaxially or basally. Staminate inflorescences more slender than pistillate inflorescences, 3-10 cm long and 1 cm wide with no indication of lateral scars on the lower part of the axis, bracts spirally ar- ranged and persistent; more flexible than the bracts of the pistillate inflorescence; bracts ~ 1 mm long, narrow proximally and slightly ex- panded distally, apex acute or rounded; ?4-6- 15 bilocular stamens per bract. Pollen grains triporate, isopolar, with con- vex sides; diameter 13.5-/5.5-25.5 ixm, pores not conspicuously elaborated, but atriate; each hemisphere bearing one or two narrow (~ 1 ixm. wide), weakly to strongly arcuate, Y-shaped, or circular pseudocolpi; surface sculpture finely granulate. Diagnosis: Platycarya strobilacea has nine advanced features not seen in other extant gen- era: 1) condensed pistillate inflorescence; 2) woody primary bracts on the pistillate inflo- rescences; 3) receptacle of the fruit Ariented abaxially; 4) fusion of the lateral sepals and bracteoles; 5) loss or reduction of the median sepals; 6) anteriorly-posteriorly flattened nut- lets; 7) spiral thickenings on vessel elements; 8) pseudocolpate pollen; 9) attenuate laminar apices. We consider features 1, 2, 4, and 8 to be characteristic of the whole genus. Platycarya americana Hickey, emend. (Fig. 3- 14,51) Salix heartensis Hickey 1977 G. S. A. Mem. 150: 117, 118; PI. 13, fig. 1,2, 5. Platycarya americana Hickey 1977 G. S. A. Mem. 150: 118, 119; PI. 14, fig. 4-8; PI. 15, fig. 2-4, 7. Lamanonia borealis Hickey 1977 G. S. A. Mem. 150: 129-131; PI. 26, fig. 9; PI. 27, fig. 2-4, 6; PI. 28, fig. 1-6; PI. 29, fig. 1-4; PI. 30, fig. 1-4. Cyperacites sp. Hickey 1977 G. S. A. Mem. 150: 113; PI. 8, fig. 4, 5. Platycarya n. sp. "A" Bebout 1977 unpub- lished doctoral dissertation, Penn. State Univ., pp. 195-197; PI. 13; fig. 8-10. Emended description: Leaves probably odd pinnately compound with more than 9 leaflets, though no whole leaves preserved; inferred from size variation and proportion of sym- metrical to asymmetrical leaflets. Leaflets sym- metrical to asymmetrical, especially at the base; straight to falcate, narrow ovate to ovate-lan- ceolate, sometimes very narrow elliptic to el- liptic, or obovate; length 4.5-19.2 cm; width 1.0-5.5 cm; length-width ratio continuously varying from 2.6-10.2; apex attenuate; base acute to decurrent, leaflet substance thickened and margin rounded to truncate at the very base of the blade where it joins the petiolule, in < 5% of the leaflets the blade completely decurrent; margin serrate, serrations mostly D4 or A4, sometimes B4 or a previously unrec- ognized type that has a reverse acuminate basal margin (concave below, convex above) with a convex apical side; leaflet texture chartaceous to subcoriaceous; petiolule long. Venation pin- nate; midvein straight or slightly curved and 394 AMERICAN JOURNAL OF BOTANY [Vol. 71 stout. Secondaries craspedodromous to semi- craspedodromous; 14-44 pairs arising from the midvein at an angle of 50? to 85?; secondary course gently and uniformly curved or abruptly upcurved at the margin, secondaries branching just before reaching the margin and sending one branch to the tooth or sinus and a generally weaker branch along the margin to the super- adjacent tooth or secondary; intersecondaries rare to common. Tertiary veins opposite per- current; course retroflexed, departing from the basal sides of the secondaries at an acute to right angle and meeting the distal side of the secondaries at a right to obtuse angle; tertiaries near the midvein recurved, sometimes forming intersecondaries of variable length. Quater- nary vein net not easily distinguished from quaternaries and higher. Areolation closed (no freely ending veinlets); areoles somewhat vari- able in size and shape, mostly 3-5 sided and less than 0.3 mm in maximum diameter. Teeth glandular Cunonioid; principal vein a second- ary or branch that enters the tooth centrally, or eccentrically on either side, and terminates at the apex; principal vein frequently gives off an almost equal branch to the sinus, in some teeth the principal vein travels directly to the sinus; in either case, the vein plexus at the sinus gives offa tertiary vein that passes up the mar- gin of the lamina to the apex of the superad- jacent tooth; this marginal branch is usually paralleled on the medial side by the distal part of a secondary, thus forming two veins that follow the leaf margin near the edge (Fig. 25). Pistillate inflorescences (1.5-4.0 cm long; 0.8-1.3 cm wide; woody bracts arranged spi- rally on a stout peduncle that is 2-3 mm broad at the base and as much as 3 cm long, and has scars below the base of the fertile portion. Bracts ovate with an acuminate beak as much as 2 mm in length; margin entire; average dimen- sions 7 mm long by 2.5 mm wide; bracts often slightly curved; venation obscure. Staminate inflorescences slender, bearing spirally arranged thin-textured bracts; inflo- rescences 2-4 mm wide and 6-12 mm long, basal part of the axis about 1.5 mm wide and lacking any lateral scars; bracts ~ 1 mm long, ovate in shape with a rounded apex, possibly constricted near the attachment to the axis; each bract bearing more than 3 but probably fewer than 7 stamens; anthers bilpcular and attached to the bract by a filament; anther loc- ules containing numerous pseudocolpate pol- len grains (described below). Fruits dorsiventrally flattened, .bilaterally symmetrical, winged nutlets with two laterally arranged locules (Fig. 7, 8); overall shape ob- ovate, length 3.5-4 mm, width approximately equal to or slightly exceeding length, recepta- cular (basal) end of the fruit containing a rounded socket; distal end of the fruit deeply embayed between the wings, median carinal sepals variable in size but sometimes extending beyond the distal end of the nutlet; surface of the nutlet marked by a longitudinal furrow (the impression of the primary partition of the nut- let) and by a pair of exmedially curved lateral furrows originating at the basal end. Pollen triporate, isopolar, with convex sides, and unelaborated, atriate pores; diameter 18.5- 23.5 nm; each hemisphere marked by one or two arcuate, circular, or Y-shaped pseudocol- pi; sculpture finely granulose. Systematic placement: Platycarya ameri- cana shares all the floral specializations of P. strobilacea except for loss of the median sepals. Minor differences in the shape of the primary bracts and the fruits allow the two species to be distinguished, but the major differences are found in the leaves. Eight leaf features of P. americana are absent or very rare in P. stro- bilacea: basilaminar tissue specialization; closed areoles; well-developed intersecondar- ies; long petiolules; decurrent or acute leaflet bases; rounded tooth apices; two parallel mar- Fig. 3-14. Platycarya americana. Key to abbreviations at the end of this caption. 3. Limonitized pistillate inflo- rescence. USNM 356409. x 2.7. 4. Compressed pistillate inflorescence showing short axis with slightly expanded base. USNM 356368. x 1.5. 5. Left, Paratype, pistillate inflorescence, showing ovate, acute-tipped, closely set bracts. USNM 43167A; Right, base of leaflet showing elaboration of basilaminar tissue (arrows). USNM 356408A. x 3.0.6. Staminate inflorescence. USNM 167478. x4.0. 7. Fruit with a partial locule cast showing the impression of the median septum (arrows). USNM 356406'. x7.0. 8. Paratype, limonitized fruit showing median septum and basal attachment scar. USNM 16748IB. x8.0. 9. Leaflet base showing paired basilaminar tissue elaboration (arrows). USNM 356371. x2.9. 10. Small mass of pollen macerated from an anther extracted from the staminate inflorescence in Fig. 12. Note one grain with an almost circular pseudocolpus (arrow). x250. 11. Pollen grain from same anther, with a Y-shaped pseudocolpus. x500. 12. Staminate inflorescence showing axis, primary bracts, and anthers. USNM 356367. x9.0. 13. Probable lateral leaflet, with elliptic shape, long petiolule, and rare lack of elaborated basilaminar tissue. USNM 356370. xl.0. 14. Probable terminal leaflet with obovate shape and specialized basilaminar tissue. USNM 356369. x0.9. Abbreviations: a, anthers; at, attachment area of fruit; ax, primary axis of inflorescence; B, primary bract; b, bracteole; lc, locule cast; Is, lateral sepal; ms, median sepal; n, nutlet body; p, median partition; sm, stigmatic surface; st, style arms. March, 1984] WING AND HICKEY?PLATYCARYA PERPLEX 395 396 AMERICAN JOURNAL OF BOTANY [Vol. 71 ginal veins; and obtuse junctions of the ter- tiaries with subadjacent secondaries. All these features are primitive for the Platycaryeae (see section on phylogeny for discussion of char- acter polarities and outgroups) and therefore do not provide a basis for placing the fossil species in a separate genus. Nevertheless, P. americana is considerably less specialized than its living congener. Occurrence: See Table 1 for localities. The two formations that have produced P. amer- icana also have yielded mammalian faunas. USNM locality 14048 in the upper part of the Camels Butte Member (Hickey, 1977) of the Golden Valley Formation, Stark Co., N.D., is stratigraphically above a mammalian fauna that has been identified as mid-Wasatchian (D. M. Schankler, written commun., 1980). U.S. Geo- logical Survey (USGS) locality 9394 is near the top of the Wasatch Formation on Pumpkin Buttes, Wyoming, and is stratigraphically higher than mammalian faunas determined to be "Lostcabinian" (Delson, 1971). Thus, mammalian biostratigraphic correlation places both localities as later early Eocene or younger. In the areas of the P. americana localities, both the Wasatch and the Golden Valley Forma- tions are overlain unconformably by Oligocene White River Formation. Collection: Holotype: pistillate inflores- cence?USNM 42992. Paratypes: pistillate in- florescences ? Princeton University (PU) 20060, 20061, 20063, and USNM 43165, 43166, 43167A and B, 43171; fruits-USNM 167481A and B, 167482, 167515, 167516. Referred material: staminate inflorescences? USNM 43172A and B, 167478, 167479 (fig- ured specimens of Cyperacites sp.), USNM 356366A and B, 356367; fruits-USNM 356365; leaflets-USNM 356363A and B, 356362, 356364, 356408A and B, 356361A and B, 356369, 356370, 356371 and speci- mens listed under the original description of Lamanonia borealis, including the holotype, USNM 43026. Platycarya castaneopsis (Lesquereux) Wing et Hickey, comb. nov. (Fig. 15-27, 51) Quercus castaneopsis Lesquereux 1883 U.S.G.S. Rept. Terr. 8: 155; PI. 28, fig. 10. Knowlton 1923 U.S.G.S. Prof. Pap. 131:161 (synonymy not accepted). Berry 1932 Am. Mus. Nov. 527: 10; text fig. 2. Brown 1934 U.S.G.S. Prof. Pap. 185-C: 55 (synonymy not accepted). Brown 1937 U.S.G.S. Prof. Pap. 186-J: 172; PL 50, fig. 9. Platycarya sp. Leopold and MacGinitie 1972 Development and affinities of Tertiary floras in the Rocky Mountains 177-179; PL I, fig. 9. Platycarya sp. Nichols and Ott 1978 Paly- nology 2:110, PL 2, fig. 14. Emended description: Leaves probably odd pinnately compound, though no whole leaves preserved; inferred from degree of asymmetry, size and shape variation of leaflets. Leaflets symmetrical to strongly asymmetrical, wide to very narrow elliptic, ovate to lanceolate, or narrow obovate; obovate leaflets always sym- metrical; length 5-20 cm, width 2.2-8.0 cm; apex usually attenuate, very rarely acute or obtuse; base decurrent; margin cut by evenly distributed, small to medium serrations (A3, A4, B3, B4, and the "reverse acuminate" type described above); margin sometimes enrolled, obscuring teeth; texture chartaceous; petiolule long. Venation pinnate; primary vein moder- ately thick, usually somewhat curved. Second- aries craspedodromous to semicraspedodrom- ous; 9-20 pairs arising from the midvein at about 90?, turning sharply upward near their terminations to follow the margin; intersecon- daries usually common. Tertiaries opposite percurrent; course retroflexed, origin on subad- Fig. 15-27. Platycarya castaneopsis. 15. Compression of pistillate inflorescence showing widely spaced bracts and lateral scars (arrows) at presumed attachment points for lateral staminate inflorescences. UCMP 9004. x 1.1.16. Pistillate inflorescence bearing an unshed winged fruit (arrow). UCMP 9003. xl.4. 17. Dispersed fruit showing long median sepal fused to lateral sepal. USNM 356402. x4.9. 18. Pistillate inflorescence, note squarish bracts with blunt apices. USNM 356407. x 1.7.19. Lateral style arms and terminal stigmatic surfaces. USNM 356401. x 15.0. 20. Fruit showing fusion of lateral sepal and bracteole, note shallow attachment scar at fruit base. USNM 356409. x6.2. 21. Close-up of counterpart of specimen in Fig. 16 showing fruit attached to pistillate inflorescence, note lateral sepal, bracteole, and nutlet body. UCMP 9003. x 9.3. 22. Less developed fruit than in Fig. 21, note bracteoles, and styles. UCMP 9003. x5.8. 23. Probable basal lateral leaflet, note asymmetry and low length to width ratio. USNM 356404. xl.4. 24. Probable lateral leaflet, note attenuate apex and decurrent base. USNM 356405. x 1.0. 25. Venation, note secondaries curving abruptly upward to parallel margin, obtuse junction between retroflexed tertiaries and subadjacent secondaries, double ascending branch in the marginal venation, and closed areolation. USNM 324549. x 4.3. 26. Probable terminal leaflet, note long petiolule with expanded base, obovate shape, long acute base, and inter-secondary veins. USNM 324500. x 1.1.27. Dispersed pollen grain from locality WhBg, at which P. castaneopsis comprises ~ 50% of the megaflora and this pollen type makes up over 30% of the palynoflora. x 500. Abbreviations as in Fig. 3-14. March, 1984] WING AND HICKEY?PLATYCARYA PERPLEX 397 "*_-"%#& .. 398 AMERICAN JOURNAL OF BOTANY [Vol. 71 jacent secondary often obtuse, medial tertiaries recurved to midvein, forming intersecondar- ies. Quaternaries not easily distinguished from higher order veins. Areolation very strongly impressed, but not oriented; areoles 0.15-0.30 mm in maximum diameter, usually 3-5 sided and lacking freely ending veinlets. Tooth ve- nation as in P. americana. Pistillate inflorescences 1.5-4.5 cm long, 0.5- 0.9 cm wide; woody bracts arranged spirally on a slender woody axis with lateral scars be- low. Bracts wide ovate and entire with a round- ed or slightly acuminate apex, 2-6 mm long and 2-4 mm wide; bracts more widely spaced than in P. strobilacea or P. americana, ~2 mm apart; bract venation consisting of ~ 9 parallel strands, the three central strands thicker than the rest. Staminate inflorescences not re- covered. Fruits dorsiventrally flattened, bilaterally symmetrical winged nutlets with two locules divided by a median partition; 3-9 mm across (lateral direction), 2-4 mm in the proximal- distal directon; style arms <0.5 mm long, transverse and carinal, not united at base, stig- matic surfaces on admedial and terminal sur- faces of the stylar arms; ovary inferior, perianth consisting of four persistent sepals and two bracteoles; median sepals roughly triangular in shape, always continuing beyond the stigmatic area, usually 1-2 mm long; abaxial midline of the lateral sepals fused to the bracteoles along V2 to 4/5ths of their distal edge, thus forming a trough-like structure at the distal end of the fruit (Fig. 50); bracteoles 3-5 mm long and 1 mm across, directed laterally and distally from the wall of the nutlet, each bracteole with sev- eral thin vascular strands that are separate to the rounded apex, where they form a plexus; base of the nutlet bearing a small mound of sediment or a shallow embayment apparently representing the sediment-filled area of attach- ment, or "socket," this socket with an abme- dial-basal orientation. Pollen triporate, isopolar, with convex sides, and unelaborated, atriate pores; diameter 18.5? 23.5 jum; each hemisphere marked by one or two arcuate pseudocolpi; sculpture finely gran- ulose. Systematic placement: Platycarya casta- neopsis is less similar to P. strobilacea than is P. americana because it has more generalized fruits and inflorescences. P. castaneopsis has four plesiomorphies not seen in the other two species: well-developed median and lateral se- pals, partially terminal stigmatic surfaces, a more basally oriented receptacle, and a less- compressed pistillate inflorescence. The rounded bracts of the pistillate inflorescence may also be a primitive feature. The leaves of P. castaneopsis are almost identical with those of P. americana except that they usually lack the expanded basilaminar tissue. We place this fossil species in Platycarya because it has six of the diagnostic features of the genus. How- ever, it is clearly more generalized than either P. strobilacea or P. americana. Occurrence: See Table 1 for localities. Many of the sites that produce P. castaneopsis can be correlated with sequences of vertebrate fau- nas. P. castaneopsis from the Niland Tongue of the Wasatch Formation in the ?Washakie Basin occurs in the part of the section that contains "Lostcabinian" mammal faunas (Roehler, 1969). In the Great Divide Basin P. castaneopsis occurs in the Niland Tongue of the Wasatch Formation, which is probably late early Eocene in age (Piperingos, 1961). In the Wind River Basin, megafossils of P. cas- taneopsis are abundant in the upper parts of the Wind River Formation, including near the type area of the Lost Cabin Member (USGS locality 9051), and in volcaniclastic rocks on the western side of the basin. The first occur- rence of P. castaneopsis megafossils in the Big- horn Basin coincides almost exactly with the appearance of "Lostcabinian" mammal faunas (Schankler, 1980; Wing, 1981). The type spec- imen off. castaneopsis is from the Green River Formation in Uinta County, Wyoming, prob- ably from the so-called fish-bearing beds on Fossil Butte. These rocks are also of late early Eocene age (Gazin, 1959). Thus the known occurrences of the species are all late early Eocene. Platycarya americana and P. castaneopsis are easily distinguished by characters of all megafossil organs, but the recognition of two species is reinforced by their geographic dis- tribution. P. americana is known only at lo- calities east of the Front Range of the Rocky Mountains, while P. castaneopsis occurs ex- clusively west of the Front Range (Fig. 1). Within the limits of resolution of mammalian biostratigraphy the two species are contem- poraneous. To our knowledge these are the first reported contemporaneous allopatric species in the Tertiary plant record. Collection: Lectotype: leaflet?USNM 1575. Referred material: leaflets-USNM 324500, 324548, 324549, 356404; pistillate inflores- cences-USNM324502AandB,UCMP9003, 9004; fruits-USNM 356401, 356402,356409. Platycarya richardsonii (Bowerbank) Chandler (Fig. 33, 34) Petrophiloides richardsonii Bowerbank 1840 Fossil fruits of the London Clay, p. 43-47; PL March, 1984] WING AND HICKEY?PLATYCARYA PERPLEX 399 st Jg&fciit ? \SmX Fig. 28-35. Fossil platycaryoids. 28-29, Platycarya manchesterii sp. nov. 28. Holotype. Note short basal portion of axis. USNM 356372A. xl.4. 29. Inflorescence showing rounded bracts. USNM 356373A. x2.1. 30-32, Hooleya lata sp. nov. 30. Holotype. Note elongated, lateral style arms, large median and lateral sepals, partial fusion of the lateral sepals to the large bracteoles, and the shallow embayment at the base of the nutlet. There is no indication that the bracteoles are fused to one another near the nutlet base, as is observed in Pterocarya spp. USNM 356376. x 15. 31. Distal end of nutlet. Note elongated, lateral arms borne on a common style. Above the style arms are the two, overlapping median sepals. USNM 356383A. x30. 32. Unusual fruit with highly divided perianth. USNM 356381. x5.4. 33-34, Platycarya richardsonii. 33. Locule cast showing median partition. USNM 356410. xlO. 34. Same specimen showing basal orientation of nutlet attachment. USNM 356410. x5. 35. Central portion of Hooleya lata fruit showing median and lateral sepals, x 15. Abbreviations as in Fig. 3-14. 9, fig. 9-15; PI. 10, fig. 5-8. Reid and Chandler 1933 The London Clay flora, pp. 133-140; text fig. la-h; PL II, fig. 7-20 (synonymy accepted); Chandler 1961 The lower Tertiary floras of southern England I, p. 136; PL 14, fig. 11-14. Platycarya richardsonii (Bowerbank) Chan- 400 AMERICAN JOURNAL OF BOTANY [Vol. 71 dler 1964 The lower Tertiary floras of southern England IV, p. 110, 111. Emended description: Comparisons of Platycarya species are presented in Table 2. Reid and Chandler (1933) noted the persistent perianth in P. richardsonii; we add it is much longer than in P. strobilacea (see Reid and Chandler, 1933, PL II, fig. 12, 13). We also note the tendency for the attachment of the nutlet to be directed more basally (Fig. 46; Reid and Chandler, 1933, PI. II, fig. 12; text fig. li) than is typical in mature inflorescences of Pla- tycarya strobilacea. P. richardsonii also has shorter, broader bracts than does P. strobila- cea. In these features P. richardsonii more closely resembles P. americana and P. casta- neopsis than it does P. strobilacea. Systematic placement: We retain P. richard- sonii in Platycarya because it has four of the diagnostic autapomorphies of the genus: con- densed pistillate inflorescences, woody pri- mary bracts, anteriorly-posteriorly flattened nutlets, and spirally thickened vessel elements. Pseudocolpate pollen assigned to Platycarya has also been recovered from the London Clay, although no established association exists be- tween pollen and pistillate inflorescences. In spite of the prevalence of mosaic evolution, the unknown organs of P. richardsonii prob- ably would not present features that would re- move the species from Platycarya. Platycarya bognorensis (Chandler) Wing et Hickey, comb. nov. Pterocaryopsis bognorensis Chandler 1961 The Lower Tertiary floras of southern England I, pp. 142, 143; PI. 17, fig. 1, 2. Pterocaryopsis elliptica Chandler 1978 Sup- plement to the lower Tertiary floras of southern England, Pt. 5, pp. 19, 20, PI. 3, fig. 4, 5. Systematic placement: In her original de- scription of Pterocaryopsis bognorensis, Chan- dler (1961) commented on the deep excavation of the lower surface near the point of attach- ment and suggested that the fruit must have been borne in a loose spike or cluster. We con- cur with Chandler's suggestion, and note that the fruit also has wings that are perfectly par- allel to its proximal-distal axis. These features are characteristic of Platycarya fruits rather than those of the pterocaryoid line, and we therefore transfer Pterocaryopsis bognorensis to Platycarya. The single specimen t)f Ptero- caryopsis elliptica cannot be distinguished from that of P. bognorensis. Manchester (1981) suggested that Pterocar- yopsis bognorensis was the dispersed fruit of Platycarya richardsonii, although he retained the two taxa as separate entities. The mor- phology of these two taxa is consistent with his conclusion, but in the absence of any mor- phological or statistical association between the two forms we too leave them in separate species. Platycarya manchesterii Wing et Hickey, sp. nov. (Fig. 28, 29) Description: Known only from pistillate in- florescences. Axes 3-6 cm long and 1-2 mm wide, bearing a spirally arranged series of closely overlapping woody bracts along the distal Vi to %ths of the axis. Bracts ovate, acute-tipped, 1.5-3.0 mm wide at the base and 2-4 mm long. Base of the axis slightly expanded and termi- nating in a sharply defined concave surface. One small scar present on the axis just below the fertile part. Systematic placement: We have placed P. manchesterii in Platycarya because it has two of the diagnostic features of the genus: woody primary bracts and a condensed pistillate in- Fig. 36-47. Living Juglandaceae. 36. Cleared leaflet of Engelhardia roxburghiana Wall., showing decurrent base and obtuse tertiary-secondary junctions. USGS Cleared Leaf Collection 891. x 1.0. 37-42, Platycarya strobilacea Sieb. et Zucc. 37. Mature fruit with abaxially elongated attachment and reduced lateral sepals. University of California Herbarium (UCH) M286369. x6.3. 38. Primary bract from mature pistillate inflorescence, note elongate shape and short peduncle (arrow) to which the fruit is attached. UCH M286369. x6.3. 39. Cleared bract of immature pistillate inflorescence, note three primary vascular strands in the central portion of the bract (1,2, 3) and separate vascularization of the small lateral lobes (L). x 30. 40. Cleared bract and anthers from staminate inflorescence, note three primary vascular strands in central portion of bract, x 30. 41. Cleared immature fruit, note vascular strands passing to the area of the reduced median sepal (arrows), x 35.42. Cleared immature fruit, note that the lateral appendages are vascularized from two sources, the lateral bundles (lb) supplying the bracteoles and the median bundles (mb) supplying the median and lateral sepals. x35. 43. Cleared leaflet of Alfaroa manningii Leon showing obtuse secondary-tertiary junctions, frequent intersecondary veins, and closed areolation. USGS Cleared Leaf Collection 889. x 1.0. 44-45, Oreomunnea mexicana (Standl.) Leroy. 44. Cleared leaflet showing elaboration of basilaminar tissue ("auriculae" of Stone), this region of the leaflet is densely pubescent; also note closed areolation. USGS Cleared Leaf Collection 897A. x6.4. 45. Cleared leaflet showing secondaries almost perpendicular to midvein, and frequent intersecondaries. USGS Cleared Leaf Collection 897A. xl.0. 46, 47, Platycarya strobilacea. 46. Aberrant leaflet with closed areolation, small teeth, and a strong double ascending branch in the marginal venation. UCMP Cleared Leaf Collection 541. x 10. 47. Normal leaflet with orthogonal secondary-tertiary junctions and areoles with freely ending veinlets, also note slight elaboration of basilaminar tissue. UCMP Cleared Leaf Collection 541. x2.1. Abbreviations as in Fig. 3-14. March, 1984] WING AND HICKEY?PLATYCARYA PERPLEX 401 402 AMERICAN JOURNAL OF BOTANY [Vol. 71 florescence. However, this species occurs at the same locality as dispersed platycaryoid fruits (described below) that belong in Hooleya. If these dispersed organs prove to belong to the same species, then P. manchesterii must be transferred to Hooleya because of the apo- morphies, such as elongate style arms, seen in Hooleya. (See discussion under H. lata.) Living and fossil pistillate inflorescences of Platycarya are quite similar in overall ap- pearance and organization, the main distinc- tions being in the shape of the inflorescences, the shape of the individual bracts, and in the amount of separation between successive bracts. This species resembles P. castaneopsis in having short, rounded bracts (4 mm), but it is similar to P. americana in having a short basal part of the axis that lacks bracts, and in having closely set bracts. The compressed pis- tillate inflorescences provide few characters that help to resolve relationships within the Platy- caryeae, yet P. manchesterii does have a unique combination of overall shape, bract outline and spacing that distinguishes it from other species of Platycarya. The specific epithet is given in honor of Ste- ven Manchester in recognition of his contri- butions to studies of fossil Juglandaceae, and his valuable assistance with this study. Occurrence: USGS locality 8637 in the lower Clarno Formation near Mitchell, Oregon. The age of these sediments is probably late early to early middle Eocene (J. A. Wolfe, pers. com- mun, 1982). Collection: Holotype: USNM 356372A and B. Referred material: USNM 356373A and B, 356374A and B, 356375A and B. Germs Hooleya Reid et Chandler 1926 Emended description: Known only from fruits. Specific name in parentheses denotes conditions observed only in that species. In- ferior, posteriorly-anteriorly flattened (H. Her- mis), winged nutlets with a rounded outline and two probably transverse locules (H. lata). Perianth usually four parted (H. lata), median sepals free, but lateral sepals partly fused to the distal margin of the wings (H. lata), rarely perianth irregularly divided into many narrow lobes. Fruit wings variable in shape, often ovate, obovate, or rounded, with an irregular or erose margin; wing orientation parallel to the fruit axis; venation a loose system of bifurcating strands that rarely anastomose. Apex of the fruit with two elongated (1.0-1.5 mm), trans- verse style arms bearing two (H. lata), or four (H. Hermis) stigmas. Base of nutlet slightly concave, with basally oriented attachment area. Surface of the wings and nut bearing discoid papillae (H. Hermis). Overall dimensions 5- 19 mm wide by 3-7 mm high; nutlets 2-4 mm wide. ? Emended diagnosis: Hooleya is distin- guished from all other juglandaceous genera by a combination of features: nutlet anteriorly- posteriorly flattened, lateral sepals partly fused to bracteoles, style arms elongated, bracteolar wings enlarged, shallowly concave basal at- tachment. Hooleya lata Wing, sp. nov. (Fig. 3(3-32, 35, 52) Description: Bilaterally symmetrical, winged nutlets, 5-19 mm wide and 3-7 mm high (proximal-distal direction) at the midline. Nut- let cast with a faint midline depression (trans- verse locules?); outline rounded, 2.5-4.0 mm wide at the broadest point and shallowly con- cave at the base; attachment area oriented ba- sally. Distal end of the nutlet bearing two elon- gate (1.0 mm), transversely oriented stylar arms. Lateral wings of the fruits unusually variable in shape, obovate with rounded tips in the larger specimens, ovate with obtuse tips in small specimens. Lateral wings extending 2-7 mm from the nutlet wall, 3-7 mm across at the greatest basal-distal extent. Wing venation ob- scure, but appearing to consist of thick strands making a series of low angle dichotomies in passing from the nutlet to the wing margin. Wing margin usually irregular or erose. Peri- anth of four parts: two lateral and two median sepals. Laterals 0.5-3.0 mm long, free at the tip but fused for part of their length to the medial part of the distal margin of the wing. In some specimens each lateral sepal can be seen to be a double structure with an abaxial and adaxial side that come together at the distal wing margin to form a trough-like structure, the bottom of the trough being at the distal wing margin. Median sepals elliptic and 1-3 mm long, arising from near the middle of the nutlet and extending well beyond the stigmatic area. In some specimens both sepals of the pair can be seen because they do not completely overlap (Fig. 31, 52). Systematic placement: We interpret these fossil fruits as having a floral envelope of two bracteoles that form most of the lateral wing tissue, two lateral sepals that are partly fused to the bracteoles at their own abaxial midline, and two largely free median sepals. Although the overall fruit shape of H. lata is similar to that in some Pterocarya species, this similarity results from a combination of shared primitive features (four distinct sepals, rounded nutlet March, 1984] WING AND HICKEY?PLATYCARYA PERPLEX 403 Juglans Pterocarya and Carya Engelhardia Oreomunnea Alfaroa ^V )r Hooleya \. \^ / Hooleya Hermis >w ^^ / lata / Platycarya strobilacea ^V ^V s \^ / Platycarya Platycarya Platycarya S castaneopsis richardsonii americana Fig. 48. Relationships in the Juglandaceae. This is one of two equally parsimonious trees based on a cladistic analysis of 30 characters of the leaves, fruits, inflorescences, and pollen (Wing and Hickey, unpublished data). See text for more detailed explanation. outline, basal attachment), and parallel adap- tations for wind dispersal (large bracteolar wings). H. lata lacks three apomorphies seen in all Pterocarya fruits: median style arms, pri- mary bract borne on the fruit, and fusion of the bracteoles to one another near the fruit base. (This last feature causes the fruit wings to be oriented at an angle to the axis of the nutlet and is readily visible in fossil material, e.g., Pterocarya protostenoptera Tanai in Ish- ida, 1970, PI. 5, fig. 4, 7.) H. lata has an extremely generalized fruit morphology, but three factors favor its inclu- sion in the Platycaryeae. One, it has fused lat- eral sepals and bracteoles, a synapomorphy with Platycarya. Two, the features in which H. lata differs most from P. strobilacea are found in intermediate states in P. castaneopsis and P. americana. Three, H. lata occurs at the same locality as the pistillate inflorescence P. man- chesterii, and they may belong to the same biological species. H. lata (and H. Hermis) have elongate style arms, an advanced feature (Manning, 1940) not seen in any species of Platycarya. The great enlargement of the bracteoles and median se- pals may also be specialization defining a sep- arate genus. Furthermore, Hooleya lacks all but one of the apomorphies of Platycarya. This combination of primitive and derived states is sufficient to establish Hooleya as a distinct ge- nus in the Platycaryeae. Occurrence: Lower part of the Clarno For- mation near Mitchell, Oregon; USGS locality 8637. Collection: Holotype: USNM 356376. Re- ferred material: USNM 356377A and B, 356383A and B, USNM 356388A and B, and USNM 356378-356390. DISCUSSION?Phytogeny? The species re- viewed and described above greatly increase the known range of morphologies in the Platy- caryeae, and thus offer a broader base both for interpreting the evolution of the tribe and as- sessing its relationship to the rest of the family. These problems have been approached through a cladistic analysis of the Juglandaceae that uses more than 30 characters and that treats the Rhoipteleaceae, Myricaceae and Fagaceae as outgroups; the character state polarities and branching sequences discussed below are from that work (Wing and Hickey, unpubl. data). A summary diagram is given in Fig. 48. Recent hypotheses on the relationships of the Platycaryeae have treated the tribe as an early offshoot of the Juglandaceae, implying that it is a sister group to the rest of the family (e.g., Stone and Broome, 1975; Manning, 1978). Our phylogenetic reconstruction differs by 404 AMERICAN JOURNAL OF BOTANY [Vol. 71 TABLE 2. Summary of character differences between species o/Platycarya. FEVS = freely ending veinlets Characters P. strobilacea P. castaneopsis P. americana P. richardsonii 1? bract of pistillate in- elongate with ovate with round- elongate to ovate ovate, acuminate florescence acute apex ed apex with acute apex apex Bract spacing close lax close close 1? bract of staminate strap shaped ? ovate with round- ? inflorescence acuminate apex ed apex Bracteoles very short very long moderate moderate? Median sepals reduced or ab- sent very long moderate moderate? Receptacle orientation abaxial basal/abaxial basal/abaxial basal/abaxial Stigmatic surface adaxial adaxial/terminal ? ? Pollen size 13-18 pm 17-24 nm 17-24 Mm ? Basilaminar region frequently with hairs unelaborated elaborated ? Petiolule short long long 1 Leaflet base rounded or cor- date decurrent decurrent ? Intersecondary veins weak and few strong and many strong and many ? Tertiary junction with orthogonal obtuse obtuse ? subadjacent 2? Tertiary course straight retroflexed retroflexed ? Areolation >20?/o w/FEVS < 10% w/FEVS < 10% w/FEVS ? placing Platycaryeae and Engelhardieae as sis- ter tribes that together are the sister group of the rest of the family. Synapomorphies in leaf- let architecture between P. americana, P. cas- taneopsis and the Engelhardieae are the pri- mary data for this relationship. Seven foliar character states are shared by Platycarya and some or all of the Engelhar- dieae: 1) Basilaminar region with dense hair patches (Alfaroa mexicana [Stone, 1968] and P. strobilacea [Kuang and Lu, 1979; and pers. observ.]) or expanded tissue (Oreomunnea mexicana [Fig. 44, 45; Stone, 1972], P. amer- icana [Fig. 5, 9, 14], and rarely P. strobilacea [Fig. 47]); 2) Fewer than 10% of areoles with freely ending veinlets (Oreomunnea [Fig. 44- 46; Wolfe, 1959], Alfaroa [Fig. 43; Wolfe, 1959], most Engelhardia species [Wolfe, 1959; and pers. observ.], P. castaneopsis [Fig. 25], P. americana, and rarely P. strobilacea [Fig. 46]); 3) Well developed intersecondary veins (Oreo- munnea [Fig. 45], Alfaroa [Fig. 43], some En- gelhardia species, P. americana, and P. cas- taneopsis); 4) Distinction between fourth and higher order venation weak or absent (Oreo- munnea, Alfaroa, Engelhardia, P. americana, and P. castaneopsis); 5) Obtuse junctions be- tween tertiaries and subadjacent secondaries (E. roxburghiana [Fig. 36], Alfaroa, O. pter- ocarpa, P. americana and P. castaneopsis); 6) Moderately long petiolules (juvenile foliage of O. pterocarpa, E. roxburghiana, P. castaneop- sis, and P. americana); 7) Acute to decurrent leaflet bases (E. roxburghiana, O. pterocarpa, Alfaroa spp., P. americana, and P. castaneop- sis). The first four of these character states are synapomorphies; the last three are symplesio- morphies that have been lost in all other Ju- glandaceae, but are seen in the outgroups. Our hypothesis that Platycaryeae and En- gelhardieae are sister tribes is also supported by floral morphology. Three-lobed primary bracts are one of the diagnostic features of the Engelhardieae (Manning, 1978; Crane and Manchester, 1982). Manning (1978, pp. 1068- 1069) observed that primary bracts on pistil- late and staminate inflorescences of P. stro- bilacea are occasionally three-lobed. Clearings of the primary bracts of P. strobilacea (Fig. 39, 40), display a vascular pattern very similar to that of Oreomunnea: the central portion of the bract is supplied by three major strands that pass almost directly to the apex of the lobe, and the lateral lobes are each independently supplied from the base of the bract by three veins. Comparative morphology (Manning, 1940) and the fossil record (Berry, 1916; Dilch- er, Potter and Crepet, 1976; Crane and Man- chester, 1982) suggest that the Engelhardieae evolved from an ancestor that had unlobed primary bracts. If a lobed ancestor evolved before the Engelhardieae-Platycaryeae split, March, 1984] WING AND HICKEY?PLATYCARYA PERPLEX 405 @ ? ? abaxial adaxial ? IF ? H at hypothetical cross sections 49. P. strobilacea 50. P.. castaneopsis ms 51. P. americana 52. Hooleva lata Fig. 49-52. Camera lucida drawings of four platycaryoid fruits. 49. Platycarya strobilacea, abaxial (above) and adaxial (below) sides of a specimen with small median sepals. UCH 286369. 50. Platycarya castaneopsis, showing parts of both median sepals, near sides of both lateral sepals, and fusion of median and lateral sepals. Reconstructed cross sections show the trough-like form of the lateral sepals. USNM 356403A. 51. Platycarya americana, limonitized areas indicated by shading. USNM 16748 IB. 52. Hooleya lata, showing partial fusion of lateral sepals to bracteoles, partly overlapped median sepals, partial locule casts, and irregular margin of bracteolar wing. USNM 356376. All drawings x5. Abbreviations as in Fig. 3-14. then Platycarya has subsequently reverted to an unlobed bract, though it has retained the basic venation. We think it more likely that the common ancestor of both tribes had "in- cipiently lobed" primary bracts like those of Platycarya. Another feature that may support the alli- ance of Platycarya and the Engelhardieae is the 406 AMERICAN JOURNAL OF BOTANY [Vol. 71 Hooleya lata Platycarya castaneopsis Platycarya americana b-ls Platycarya strobilacea Fig. 53-56. Floral diagrams of pistillate platycaryoid flowers. 53. Hooleya lata. 54. Platycarya castaneopsis. 55. Platycarya americana. 56. Platycarya strobilacea. Note progressive reduction and loss of the perianth. Symbols after Manning (1978): ms, median sepal; Is, lateral sepal; b, bracteole; bis, fused bracteole and lateral sepal; ms Is, fused sepals; B, primary bract. The circle containing an x indicates the position of the primary axis of the inflorescence. possession of a pedicel on which the fruit is borne. P. strobilacea does have a pedicel-like structure (Fig. 38), although it is highly reduced and fused to the primary bract. Finally, a sister- group relationship between Platycaryeae and Engelhardieae is supported by synapomor- phies in their steroid and flavonoid chemical characteristics (Niklas and Giannasi, in prep- aration). If Platycaryeae and Engelhardieae are sister tribes, then the subfamily Platycaryoi- deae (Manning, 1978) should be expanded to include both. Nine species of Platycaryeae have now been described from Tertiary megafossils, but in spite of this diversity at trie species level there is little evidence for more than one lineage in the tribe. The leaves of P. americana and P. cas- taneopsis strongly resemble those of the En- gelhardieae because they are less specialized than leaves of P. strobilacea. Similarly, the fruits and flowers of all the fossil species are more generalized than those of P. strobilacea. Man- ning (1940, 1978) stated that the basic juglan- daceous floral envelope consists of seven parts: primary bract, two bracteoles, and four sepals. P. strobilacea usually lacks the median pair of sepals, and has highly fused lateral sepals and bracteoles (Fig. 37, 41, 42, 49, 56). P. ameri- cana and P. richardsonii have median sepals and larger bracteoles. P. castaneopsis has fused but distinguishable lateral sepals, and even larger median sepals and bracteoles. Hooleya lata has very large median sepals and brac- teoles, and the lateral sepals are only partially fused to the bracteoles. The fertile and foliar characteristics of all fossil Platycaryeae are more generalized than those of P. strobilacea in every respect, with the possible exception of elongated style arms in Hooleya. Although H. lata, P. castaneopsis, P. americana, and P. strobilacea are not an ancestor-descendent sequence (the two fossil Platycarya species are contemporaneous, and H. lata, the most primitive fruit type, is also the youngest fossil), they show that the early Eocene diversification of the Platycaryeae pro- duced little morphological innovation. The major evolutionary trends in the tribe have been shortening of the petiolules and leaflets, decreasing organization in leaflet venation, greater condensation of the pistillate inflores- cences, fusion and loss of floral envelope com- ponents, and reduction in overall fruit size (2 cm-4 mm). Paleoecology and evolution?Fossil remains of Platycarya are ubiquitous and abundant in early Eocene sediments of the Northern Hemi- sphere. Platycaryoid pollen first occurs at or just below the Paleocene-Eocene boundary, then becomes abundant in the later early Eocene. It is an important component ofpaly- nofloras throughout most of North America, including the Arctic, and Western Europe (Hail and Leopold, 1960; Hopkins, 1967; Rouse, Hopkins and Piel, 1971; Leopold and MacGinitie, 1972; Tschudy, 1973; Nichols, 1973; Elsik, 1974; Elsik and Dilcher, 1974; Kedves and Stanley, 1976; Newman, 1977; Bebout, 1977, 1980; Frederiksen and Chris- topher, 1978; Nichols and Ott, 1978; Freder- iksen, 1979, 1980; Christopher et al., 1980; Gruas-Cavagnetto, Laurain and Meyer, 1980a, 1980b; Nichols, in press; Hickey et al., 1983). Megafossils of Platycarya are abundant and numerically dominant at localities we have col- lected. Of his original discovery of Platycarya (Petrophiloides) richardsonii, Richardson March, 1984] WING AND HICKEY?PLATYCARYA PERPLEX 407 (quoted in Chandler, 1961) wrote that the spec- imens were found by the "bushel." There is little doubt that where Platycarya grew it was extremely abundant, and that it could form nearly monospecific assemblages. The small, numerous, winged nutlets of Platycarya (living and fossil) are typical of wind- dispersed, weedy plants that reproduce prolif- ically, mature rapidly, and tend to colonize relatively open areas where low light levels do not limit growth (Harper, Lovell and Moore, 1970; Stebbins, 1971; Stone, 1973). Givnish (1978) has commented that pinnately com- pound leaves provide early successional species with a metabolically inexpensive ultimate branch equivalent that can be grown quickly and then dropped during periods of physio- logical stress. The wood of Platycarya is con- sidered advanced because it has short, rela- tively wide, spirally thickened vessel elements, and vessel perforations that are exclusively po- rous (Heimsch and Wetmore, 1939). Reid and Chandler (1933) observed spiral thickenings in P. richardsonii. These features of vascular anatomy may also be adaptations to rapid growth in a stressful environment. Thus, many of the morphological features of the fossil platycaryoids are similar to specializations seen in modern plants that occupy early succes- sional or open habitats. Platycarya does not always occur with the same species in the localities where we col- lected (Appendix Table 2), however a variety of presumed herbaceous plants are regular as- sociates: Zingiberopsis isonervosa, Equisetum sp., and several ferns. In the Willwood For- mation, a species of Alnus is the dicotyledon that most often occurs with P. castaneopsis. At USGS 9394, P. americana megafossils occur in a monospecific assemblage near the top of a lens inferred to be a filled oxbow pond de- posit, suggesting that this species was dominant in the late phases of a hydrosere. These asso- ciated species and the distribution of Platy- carya fossils in the sediment also suggest that early Tertiary Platycarya grew in open, early successional habitats. The ecological requirements of the living P. strobilacea provide a final line of evidence about those of the fossil species. P. strobilacea is a shrub or small tree (Manning, 1978), and it is a major component of secondary growth in both the Mixed Mesophytic and the Sclero- phyllous Evergreen Forests of China (Wang, 1961). These separate lines of evidence yield similar conclusions on the probable ecology of fossil Platycarya. These early Tertiary species are inferred to have been small, probably shrubby trees that rapidly colonized open or unstable ground, were fast growing and maturing plants that produced large quantities of easily dis- persed winged fruits, and may have formed thickets of dense, nearly monospecific growth. This syndrome of weedy features is common to several amentiferous families such as Myr- icaceae (Elias, 1971) and Betulaceae (Cron- quist, 1981), and may well be the primitive condition of the Amentiferae in general (Stone, 1973). In contrast to the growth form of the plant itself, the phylogenetic tree of Platycarya is quite spare. This lack of morphological diver- sity may seem all the more surprising in a group that was so common and widely distributed, however we believe that the low diversity of the platycaryoid line may result from the eco- logical characteristics that permitted its "suc- cess." Early successional or weedy species take ad- vantage of ephemeral habitats by being easily dispersed and by growing and maturing rap- idly. The chief challenges of these environ- ments are physical: for example, dessication, low nutrient soils, fire, and flooding. Although the distribution of such habitats may be un- predictable in the short term, the challenges that they create are much the same over long periods of geological time. Thus early succes- sional species tolerate a wide range of physical conditions. Once a lineage becomes adapted to such an environment there is probably little selection for change, unless it is for a more rapid life cycle or greater reduction and sim- plification of the soma. A lineage such as the Platycaryeae also may produce few daughter lines because of the dif- ficulty of achieving reproductive or ecological separation between incipient species. Vrba (1980) offered such an "effect hypothesis" as one possible cause of evolutionary trends with- in higher taxa. She proposed that such trends may be the effect of differential rates of spe- ciation rather than of the selective advantage conferred by a particular trait. A lineage that speciates slowly leaves fewer descendant species than one that speciates rapidly, and thus the characteristics of the rapidly speciating lineage become more common through time, although the "adaptiveness" of those characteristics may not be causally related to the rapid speciation of the lineage. Vrba suggests that lineages with narrow environmental tolerances (stenotopes) have higher speciation rates than those that are eurytopic. In the present day, the animal-dispersed Ju- glandaceae (Juglans, Carya, Alfaroa) are more diverse than the wind-dispersed genera (Pter- 408 AMERICAN JOURNAL OF BOTANY [Vol. 71 ocarya, Oreomunnea, Engelhardia, Platycar- ya), but in the early Tertiary the reverse was the case. This change in dominance might well be attributed to the coevolution of angiosperms and mammals through the Cenozoic (Tiffney, in press), but the effect hypothesis offers an alternative explanation. The Platycarya lineage displays reduction and simplification, it is morphologically con- servative, and it has produced few if any daugh- ter lineages. Using Platycarya as a model, the overall trend in the Juglandaceae toward an- imal dispersal may result from slow speciation rates in the wind-dispersed lineages (a conse- quence of their r-selected life histories) rather than an adaptive advantage possessed by the animal-dispersed lines. LITERATURE CITED BEBOUT, J. W. 1977. Palynology of the Paleocene-Eocene Golden Valley Formation of western North Dakota. Unpublished Ph.D. thesis, Pennsylvania State Uni- versity, State College. : . 1980. Observed stratigraphic distribution of spores, pollen, and incertae sedis palynomorphs in the Tertiary section of the Cost No. B-2 well, Baltimore Canyon, Atlantic outer continental shelf. Palynology 4: 181-196. BERRY, E. W. 1916. The lower Eocene floras of south- eastern North America. U.S.G.S. Prof. Pap. 91. . 1932. Eocene plants from Wyoming. Amer. Mus. Nov. 527: 1-13. BOWERBANK, J. S. 1840. History of the fossil fruits and seeds of the London Clay. London. BROWN, R. W. 1934. The recognizable species of the Green River flora, a. U.S.G.S. Prof. Pap. 185-C: 45- 68. . 1937. Additions to some fossil floras of the west. U.S.G.S. Prof. Pap. 186-J: 163-186. CHANDLER, M. E. J. 1961. The lower Tertiary floras of southern England. I. British Museum (Natural His- tory), London. ?. 1964. The lower Tertiary floras of southern En- gland. IV. British Museum (Natural History), London. 1978. Supplement to the lower Tertiary floras of southern England. Pt. 5. Tertiary Research Group, London. CHRISTOPHER, R. A., D. C. PROWELL, J. REINHARDT, AND H. W. MARKEWICH. 1980. The stratigraphic and structural significance of Paleocene pollen from Warm Springs, Georgia. Palynology 4: 105-124. CRANE, P. R., AND S. R. MANCHESTER. 1982. Casholdia, an extinct juglandaceous fruit from the upper Palae- ocene of southern England. Bot. J. Linn. Soc. London 85(2): 89-101. CRONQUIST, A. 1981. An integrated system of classifi- cation of flowering plants. Columbia University Press, New York. DELSON, E. 1971. Fossil mammals of the early Wasatch- ian Powder River local fauna, Eocene of northeast Wyoming. Bull. Amer. Mus. Nat. Hist. 146(4): 309- 364. DILCHER, D. L., F. W. POTTER, AND W. L. CREPET. 1976. Investigations of angiosperms from the Eocene of North America: juglandaceous winged fruits. Amer. J. Bot. 63: 532-544. ELI AS, T. S. 1971. The genera of the Myricaceae in the southeastern United States. J. Arnold Arbor. Harv. Univ. 52: 305-318. ELSIK, W. C. 1974. Characteristic Eocene palynomorphs in the Gulf Coast, U.S.A. Palaeontographica, Abt. B 149:90-111. , AND D. L. DILCHER. 1974. Palynology and age of the clays exposed in Lawrence clay pit, Henry Coun- ty, Tennessee. Palaeontographica, Abt. B 146: 65-87. FREDERIKSEN, N. O. 1979. Paleogene sporomorph bio- stratigraphy, north-eastern Virginia. Palynology 3:129- 167. . 1980. Paleogene sporomorphs frgm South Car- olina and quantitative correlations with the Gulf Coast. Palynology 4: 125-179. , AND R. A. CHRISTOPHER. 1978. Taxonomy and biostratigraphy of late Cretaceous and Paleogene tria- triate pollen from South Carolina. Palynology 2: 113- 145. GAZIN, C. L. 1959. Paleontological exploration and dat- ing of the early Tertiary deposits in basins adjacent to the Uinta Mtns. Intermountain Assoc. Petr. Geol. Guidebook, 10th Annu. Field Conf., pp. 131-138. GIVNISH, T. J. 1978. On the adaptive significance of compound leaves, with particular reference to tropical trees. In P. B. Tomlinson and M. H. Zimmerman [eds.], Tropical trees as living systems, pp. 351-380. Cambridge University Press, London. GRUAS-CAVAGNETTO, C, M. LAURAIN, AND R. MEYER. 1980a. Un sol de mangrove fossilise dans les Lignites du Soissonais (Ypresien) a Verzenay (Marne). Geobios 13(5): 795-801. , , AND . 1980b. Paysage vegetal et position stratigraphique du sommet des Lignites du Soissonais dans la montagne de Reims. Geobios 13(6): 947-952. HAIL, W. J., AND E. B. LEOPOLD, i960. Paleocene and Eocene age of the Coalmont Formation, North Park, Colorado. U.S.G.S. Prof. Pap. 400-B: 260-261. HARPER, J. L., P. H. LOVELL, AND K. G. MOORE. 1970. The shapes and sizes of seeds. Annu. Rev. Ecol. Syst. 1: 327-356. HEIMSCH, C, AND R. H. WETMORE. 1939. The signifi- cance of wood anatomy in the taxonomy of the Juglan- daceae. Amer. J. Bot. 26: 651-660. HICKEY, L. J. 1977. Stratigraphy and paleobotany of the Golden Valley Formation (Early Tertiary) of western North Dakota. G. S. A. Mem. 150. . 1979. A revised classification of the architecture of dicotyledonous leaves. In C. R. Metcalfe and L. Chalk [eds.], Anatomy of the Dicotyledons, 2nd. ed., pp. 25-39. Clarendon Press, Oxford. , R. M. WEST, M. R. DAWSON, AND D. K. CHOI. 1983. Arctic terrestrial biota: paleomagnetic evi- dence of age disparity with mid-northern latitudes during the late Cretaceous and early Tertiary. Science 221: 1153-1156. ,ANDJ. A. WOLFE. 1975. The bases of angiosperm phylogeny: vegetative morphology. Ann. Mo. Bot. Gard. 62: 538-589. HOPKINS, W. S. 1967. Palynology of Tertiary rocks of the Whatcom Basin, southwestern British Columbia and northwestern Washington. Unpubl. Ph.D. thesis, Department of Geology, University of British Colum- bia. Hu, H. H., AND R. W. CHANEY. 1940. A Miocene flora March, 1984] WING AND HICKEY?PLATYCARYA PERPLEX 409 from Shantung Province, China. Carnegie Inst. Wash- ington Publ. 507. ISHIDA, S. 1970. The Noroshi flora of Noto Peninsula, central Japan. Memoirs Faculty Sci., Kyoto Univ., Ser. Geol. Mineral. 37(1): 1-112. KEDVES, M., AND E. A. STANLEY. 1976. Electron micro- scopical investigations of the Normapolles group and some other selected European and North American angiosperm pollen II. Pollen Spores 18(1): 105-127. KNOWLTON, F. H. 1923. Revision of the flora of the Green River Formation with descriptions of new species. U.S:G.S. Prof. Pap. 131-F: 133-176. KUANG, K.-z., AND A.-M. Lu. 1979. Juglandaceae. In K.- z. Kuang and P.-c. Li [eds.], Flora Republicae Po- pularis Sinica, Vol. 21: 6-44. LEOPOLD, E. B., AND H. D. MACGINITIE. 1972. Devel- opment and affinities of Tertiary floras in the Rocky Mountains. In A. Graham [ed.], Floristics and paleo- floristics of Asia and eastern North America, pp. 147- 200. LESQUEREUX, L. 1883. The flora ofthe Green River group; contributions to the fossil flora of the western terri- tories, Pt. 3. U.S.G.S. Terr. Rept. 8: 127-218. MANCHESTER, S. R. 1981. Fossil history ofthe Juglan- daceae. Unpubl. Ph.D. thesis, Department of Biology, Indiana University. MANNING, W. E. 1938. The morphology of the flowers of the Juglandaceae. I. The inflorescence. Amer. J. Bot. 25: 407-419. . 1940. The morphology of the flowers ofthe Ju- glandaceae. II. The pistillate flowers and fruit. Amer. J. Bot. 27: 839-852. . 1948. The morphology of the flowers of the Ju- glandaceae. III. The staminate flowers. Amer. J. Bot. 35: 606-621. . 1978. The classification within the Juglandaceae. Ann. Mo. Bot. Gard. 65(4): 1058-1087. NEWMAN, K. R. 1977. Palynologic biostratigraphy of some early Tertiary nonmarine formations in central and western Washington. Geol. Soc. Am. Abstr. with Programs 9(7): 1113. NICHOLS, D. J. 1973. North American and European species of Momipites ("Engelhardtia") and related genera. Geosci. Man 7: 103-117. . In press. Palynology ofthe Vermillion Creek coal bed and associated strata. In H. W. Roehler [ed.], Geological investigations ofthe Vermillion Creek coal bed in the Eocene Niland Tongue of the Wasatch Formation, Sweetwater County, Wyoming, U.S.G.S. Prof. Pap. -, AND H. L. OTT. 1978. Biostratigraphy and evo- lution ofthe Momipites-Caryapollenites lineage in the early Tertiary in the Wind River Basin, Wyoming. Palynology 2: 93-112. PIPERINGOS, G. N. 1961. Uranium-bearing coal in the central part ofthe Great Divide Basin. U.S.G.S. Bull. 1099-A. REID, E. M., AND M. E. J. CHANDLER. 1933. The London Clay flora. British Museum (Natural History), Lon- don. , AND . 1926. The Bembridge flora; Cata- logue of Cainozoic plants in the Department of Ge- ology, I. British Museum (Natural History), London. ROEHLER, H. W. 1969. Stratigraphy and oilshale deposits of Eocene rocks in the Washakie Basin, Wyoming. 21st Ann. Field Conf. Wyo. Geol. Assoc. Guidebook. ROUSE, G. E., W. S. HOPKINS, AND K. M. PIEL. 1971. Palynology of some late Cretaceous and early Tertiary deposits in British Columbia and adjacent Alberta. G.S.A. Special Paper 127: 213-246. SCHANKLER, D. M. 1980. Faunal zonation ofthe Will- wood Formation in the central Bighorn Basin, Wy- oming. In P. D. Gingerich [ed.], Early Cenozoic pa- leontology and stratigraphy of the Bighorn Basin, Wyoming. Univ. Mich. Pap. Paleo. 24: 99-114. SCHEIHING, M. H., AND H. W. PFEFFERKORN. 1982. Ta- phonomy of land plants in the Orinoco Delta: a model for incorporation of plant parts in clastic sediments of Pennsylvanian age of North America. Bot. Soc. Am. Misc. Ser. Publ. 162: 64. ? SIEBOLD, FRANZ VON, AND J. G. ZUCCARINI. 1843. Plan- tarum in Japonicum collectarum genera nova, notis characteristics illustrata. Abh. Math.-Phys. Cl. Ko- nigl. Bayer Akad. Wiss. 3: 717-750. STAFLEU, F. A., C. E. B. BONNER, R. MCVAUGH, R. D. MEIKLE, R. C. ROLLINS, R. ROSS, J. M. SCHOPF, G. M: SCHULZE, R. DE VILMORIN, AND E. G. VOSS. 1972. International Code of Botanical Nomenclature. Int. Assn. Plant Tax., Utrecht. , V. DEMOULIN, W. GREUTER, P. HIEPKO, R. LINCZEVSKI, R. MCVAUGH, R. D. MEIKLE, R. C. ROLLINS, R. ROSS, J. M. SCHOPF, AND E. G. VOSS. 1978. International code of botanical nomenclature. Int. Assn. Plant Tax., Utrecht. STEBBINS, G. L. 1971. Adaptive radiation of reproductive characteristics in angiosperms. II. Seeds and seedlings. Annu. Rev. Ecol. Syst. 2: 237-260. STONE, D. E. 1968. New World Juglandaceae. A new species of Alfaroa from Mexico. Amer. J. Bot. 55: 477-484. . 1972. New World Juglandaceae. III. A new per- spective of the tropical members with winged fruits. Ann. Mo. Bot. Gard. 50: 297-321. . 1973. Patterns in the evolution of Amentiferous fruits. Brittonia 25: 371-384. , AND C. R. BROOME. 1975. Juglandaceae A. Rich. ex Kunth. World Pollen Spore Flora 4: 1-35. TIFFNEY, B. H. In press. Seed size, dispersal syndromes and the rise of the angiosperms: evidence and hy- pothesis. Ann. Mo. Bot. Gard. TSCHUDY, R. H. 1973. Stratigraphic distribution of sig- nificant Eocene palynomorphs ofthe Mississippi Em- bayment. U.S.G.S. Prof. Pap. 743-B: 1-24. VRBA, E. S. 1980. Evolution, species and fossils: how does life evolve? S. Afr. J. Sci. 76: 61-84. WANG, C. W. 1961. The forests of China. Maria Moors Cabot Foundation Publ. Ser. 5. WING, S. L. 1980. Fossil floras and plant-bearing beds ofthe central Bighorn Basin. In P. D. Gingerich [ed.], Early Cenozoic paleontology and stratigraphy of the Bighorn Basin Wyoming. Univ. Mich. Pap. Paleo. 24: 119-126. . 1981. A study of paleoecology and paleobotany in the Willwood Formation (early Eocene, Wyoming). Unpublished Ph.D. thesis, Department of Biology, Yale University, New Haven, Conn. 1984. Relation of palepvegetation to geometry and cyclicity of some fluvial carbonaceous deposits. J. Sed. Pet. WOLFE, J. A. 1959. Tertiary Juglandaceae of western North America. Master's thesis, Department of Pa* leontology, University of California, Berkeley. . 1979. Paleogene floras from the Gulf of Alaska region. U.S.G.S. Prof. Pap. 997. 410 AMERICAN JOURNAL OF BOTANY [Vol. 71 APPENDIX: TABLE 1. Fig. l Locality no. 9 10 11 USNM 14048 USGS 9550 USNM USGS 9394 MBR T SL TL FL CQ 15M 15ME 12 WhBg 13 WhBgL 14 B 15 T2 16 TAT79 17 BCT 18 BCT2 19 USGS 8894 20 UCMPPA114. USGS 8912 21 UCMP PA 104 22 USGS 9538 23 USGS 9051 24 USGS 9397, USGS 5286 25 D4309 26 USGS 8637 SW V* sec. 29, T. 139 N, R. 97 W, Stark Co., ND. Upper Camels Butte Mbr., Golden Valley Fm. Western pit of the Hebron Brick Company; SW % SW %, sec. 4, T. 140 N, R. 90 W, Morton Co., ND. Lower Camels Butte Mbr., Golden Valley Fm. 125 ft. below conglomerate capping the 3rd Pumpkin Butte from the North. On south face. Campbell Co., Wyo. Upper Wasatch Formation, sec. 10, T. 51 N, R. 97 W, Bighorn Co., Wyo. Upper Willwood Formation. NE y4, SW VA, sec. 25, T. 49 N, R. 98 W, Bighorn Co., Wyo. Uppermost Willwood Fm. NW VA, SW 1/4, sec. 25, T. 49 N, R. 98 W, Bighorn Co., Wyo. Uppermost Villwood Fm. SW 1/4, SW 1/4, sec. 25, T. 49 N, R. 98 W, Bighorn Co., Wyo. Uppermost Willwood Fm. NE i/i, NE 1/4, sec. 35, T. 49 N, R. 98 W, Bighorn Co., Wyo. Uppermost Willwood Fm. NE VA, SW 1/4, sec. 31, T. 49 N, R. 97 W, Bighorn Co., Wyo. Uppermost Willwood Fm. NW 1/4, SE i/,, sec. 25, T. 49 N, R. 98 W, Bighorn Co., Wyo. Uppermost Willwood Fm. SW 1/4, NW i/,, sec. 32, T. 49 N, R. 97 W, Bighorn Co., Wyo. Uppermost Willwood Fm. NE i/,, SE 1/4, sec. 1, T. 48 N, R. 98 W, Bighorn Co., Wyo. Uppermost Willwood Fm. SW 1/4, sec. 6, T. 48 N, R. 97 W, Bighorn Co., Wyo. Uppermost Willwood Fm. NE 1/4, SE 1/4, sec. 1, T. 48 N, R. 98 W, Bighorn Co., Wyo. Uppermost Willwood Fm. SW i/t, SW 1/1, sec. 6, T. 48 N, R. 97 W, Hot Springs Co., Wyo. Lowermost Tatman Fm. SE %, SE 1/,, sec. 1, T. 48 N, R. 98 W, Hot Springs Co., Wyo. Lowermost Tatman Fm. NE 1/4, SW 1/4, sec. 35, T. 49 N, R. 97 W, Bighorn Co., Wyo. Lowermost Tatman Fm. SW 1/4, SE 1/4, sec. 35, T. 49 N, R. 97 W, Bighorn Co., Wyo. Lowermost Tatman Fm. 2 miles NW of SE corner T. 54 N, R. 104 W, Park Co., Wyo. Willwood Fm. On north side of highway, 10 miles west of Dubois, Fremont Co., Wyo. Wind River Fm. Fremont Co., Wyo. Wind River Fm. NE corner sec. 28, T. 33 N, R. 95 W, Fremont Co., Wyo. Wind River Fm. W end of N side of C. B. & Q. railroad cut at milepost 293 about 10 mi W of Lysite, Fremont Co., Wyo. Wind River Fm. SE 1/,, SW i/i, sec. 24, T. 24 N, R. 96 W, Sweetwater Co., Wyo. Niland Tongue, Wasatch Fm. Niland Tongue, Wasatch Fm., Carbon Co., Wyo. 1 mi. N of Black Butte, 5 mi. W of Mitchell, Oregon. Lower Clarno Fm. March, 1984] WING AND HICKEY?PLATYCARYA PERPLEX 411 APPENDIX: TABLE 2. Floral lists from representative lo- calities yielding Platycarya megafossils Locality Taxa present (other than Platycarya) MBR SL 15ME 15M TL USNM 14048 Allantoidiopsis erosa, Amesoneuron sp., Dalbergia.7, Typha-like mono- cot, Salvinia preauriculata, Zingi- beropsis isonervosa. Amesoneuron sp., Equisetum sp., in- det. monocot. Allantoidiopsis erosa, Lygodium kaul- fussi, Z. isonervosa, indet. magno- liid. Alnus sp. A. erosa, Alnus sp., Amesoneuron sp., Averrhoites affinis, Cercidiphyllum genetrix, indet. magnoliid, Dom- beya novi-mundi, Equisetum sp., L. kaulfussi, Platanus "nobilis," Z. iso- nervosa. Cnemidaria magna, L. kaulfussi, "Sparganium" stygium, Thelypteris iddingsii. Acrovena laevis, Betula hesterna, Car- polithes bryangosus, Equisetum magnum, Glyptostrobus europaeus, L. kaulfussi, Palmacites sp., Para- ternstroemia hyphovenosa, Peltan- dra primaeva, Porosia verrucosa, S. preauriculata, "Sparganium"sty- gium, Stillingia casca, Ternstroe- mites aureavallis, Z. isonervosa.