288 



LOWER EOCEXE FLORAS OF SOUTUEASTERN NORTH AMERICA. 



Sterculiocarpus eocexicus Bcirv, n. sp. 

 Platp LXXIV, figures 1-3. 



Description. — Large capsular fruit, appar- 

 ently dehiscent from the apex, consistmg of a 

 stout central peduncle surrounded by five 

 elliptical, broadly keeled capsules which are 

 united for nearly then- whole length. Total 

 Icngtii of fruit 6 centhueters. Diameter 6 

 centimeters. Diameter of pedmicle 1.25 centi- 

 meters. Sm-face smooth. Textm"e apparently 

 coriaceous or hgneous. 



This magnificent fruit is perfectly symmetric 

 and must have been of considerable consist- 

 ency. A considerable portion of the matrix 

 is cemented to the apex, as showm m the side 

 view of the specimen, which has prevented the 

 determination whether or not dehiscence had 

 commenced, or whether the individual capsules 

 were pointed distad or broadly rounded as they 

 are proxuuad. In dorsal view each capsule 

 forms an elongated ellipse, broadly and eveidj- 

 roimded below and apparently equally rounded 

 above, with a broad and not especially promi- 

 nent dorsal keel. Photographs of the speci- 

 men have been submitted to a number of 

 speciahsts familiar with the existmg flora of 

 the Tropics without arriving at any definite 

 decision regarding then- generic affinity, al- 

 though there was a rather general agreement 

 that the fruit was probably referable to the 

 Sterculiacere. I have compared it with all 

 the material representing this and alhed 

 families at the New York Botanical Garden 

 without being able to match it with living 

 forms, although a nmuber of modern genera 

 show similarities, for example, Reevesia Lind- 

 ley, a small Asiatic genus (cf. Reevsia thyrsoidea 

 Lindley, which has a whorl of 1-seeded tardily 

 dehiscent capsules), Abroma Linne fils, a 

 small East Indian and Australian genus, and 

 Hclicteres Linne, a cosmopolitan tropical 

 genus, which comprises about two score 

 existing species. There is also a resemblance 

 to some of the Dilleniacese, Euphorbiacese, 

 ZygophyUacesE, and the like. 



The oidy known fossil form at all similar 

 enough to Sierculiocarpus eocenicus to be con- 

 sidered as related is SezaneUa major Viguier,' 

 described from the wonderful casts of fossils 

 from the travertine of Sezanne made by the 



' Viguier, Ren^, Revue gc5n4rale de botanique, vol. 20, pp. 6-13, text 

 figure 1, pi. 5, figs. 1-7, 10, 1908. 



late Munier-Chahnas and m the collections of 

 the Sorbonne. SezancUa, which is as well 

 known as if it was a recent species, is based on 

 complete flowers and fruits, showing the ar- 

 rangement of the seeds m the capsules. It is 

 only about two-thirds the size of Sterculiocar- 

 pus, the individual capsides are shorter, less 

 fuU, and lack the keel of .Sterculiocarpus. The 

 form is referred by Viguier to the tribe Lasio- 

 petahv of the family Sterculiaceie. The traver- 

 tine of Sezanne is a fluvial deposit usually 

 correlated with the Thanetian or lower Eocene 

 marine sands of the Paris Basin. 



Occurrence. — Wilcox group, Frierson MLU, 

 and 3^ miles southeast of Naborton, De Soto 

 Parish, La. (collected by G. C. Matson). 



Collection. — V. S. National Museum. 



Sterculiocarpus sezanxelloides Berry, n.sp. 



Plate LXXII. figures 4-6. 



Description. — Fruit consisting of a whorl of 

 five coalescent capsules, forming a spherical 

 5-valved capsule, the units slightly free distad. 

 Length about 2.5 centimeters. Lateral diam- 

 eter sUghtly less than the length. Texture 

 coriaceous. Valves equilateral, elliptical, widest 

 in the middle and tapering about equally pro.xi- 

 mad and distad, with distmct keels. Deliis- 

 cence septicidal from the apex, the valves ap- 

 parently becommg widely separate and possibly 

 reflexcd. Placentte axile. Seeds numerous, 

 elliptical in outline. 



This species is based on the single, somewhat 

 compressed specimen figirred from a photo- 

 graph. iVlthough somewhat distorted, the five 

 septicidal valves can be readily made out, as 

 well as the impressions of some of the seeds on 

 the capsular walls. A drawing has been made 

 of the capsule before and after dehiscence, not 

 only to further characterize the species but to 

 elucidate the photographic illustration of the 

 type. There seems to be little doubt that the 

 present specimen represents a more or less 

 buoyant capsule of some Wilcox species of 

 Sterculiacese. It is not exactly like the fruit 

 of any modern member of the family knowai to 

 me, but it resembles several of the existmg 

 genera in certain particulars. It is much smaU- 

 er than Sterculiocarpus eocenicus Berry, the 

 valves are less strongly keeled, and the dehis- 

 cence is much more pronoimced, although this 

 mav be partly due to compression during fos- 

 silization. ^Vmong j)reviously described fossil 



