236 



LOWEK EOCEXE FLORAS OF SOCTHEASTERN XOETH AMERICA. 



ing), and p. homluc (Roxburg). The second 

 of these is cosmopolitan and its seeds float 

 miinjured Tor nmnths. Tliere are a number 

 of records of their occurrence in tlie drift on 

 the Irish and Scandinavian coasts. Robert 

 Bro\\Ti recorded a plant raised from a West 

 Indian seed washed up on the Irish coast, and 

 these features of distribution are discussed by 

 Hemsley, Schimper, Guppy, and Sernander. 



The fossd species of Civsalpinia are numer- 

 ous, numbering more than a score, besides 

 about an ecjual number uf forms of Ca-sal- 

 pmitcs. Tliey are largely repres(>nted in the 

 European Tertiary, commencing witli the 

 upper Eocene. In this country our previ- 

 ously known Eocene floras have been of a rather 

 different type, and leguminous forms have not 

 been discovered in them in great quantities. 



A few fossil forms that resemble the present 

 species are Leguminosites calpurnioides Saporta' 

 from the French Oligocene, whicli is practi- 

 cally identical with the larger leaflets of the 

 American Eocene form except that the French 

 form has a shorter petiolule. Cxsalpinifcs 

 coUigendus Saporta - from the lower Oligo- 

 cene of France is practically identical with the 

 smaller forms of the present species, Copmfera 

 relicfa Unger ^ from Radoboj in Croatia is also 

 almost exactly like the larger leaflets of the 

 present species. Other similar forms from the 

 Tertiary of Bolivia are described by Engel- 

 hardt as Plati podium potosianum ■* and Dre- 

 panocarpus franJcei' and are supposed to repre- 

 sent these two allied genera of papilionaceous 

 trees, which in the existing flora are confined 

 to the American Trojoics. Another fossil species 

 which closely resembles the larger leaflets of 

 C'sesalinnia wilcoxiana Berry is described by 

 Engelhardt ° as Cassia longifolia. It is from 

 the Tertiary of Ecuador. 



Csesalpinia wilcoxiana was apparently com- 

 mon throughout Holly Springs time. North- 

 ward it appears to have. l)cen replaced by species 

 of Mimosites. 



' Saporta, G. de, fetudps sur la v(SgStation du sud-cst de la France 4 

 l'i;po(iuo Icrtiaire, vol. 3, p. 189, pi. 7, flg. 7, 1867. 



2 .Saporta, G. do, Derai^res adjonctions k la flore fossile d'.Vi.\-en- 

 I'rovence, p. 121, pi. Ifl, fig. 24, 1889. 



" linger, Franz, Sylloge plantarura fossilium, vol. 2, p. 32, pi. U, fig. 11, 

 1SU2; Die fossile Flora von Radoboj, p. 154, ])1. 3, fig. 10, 18C9. 



< Engelhardt, Hermann, Natnrwiss. Gesell. Isis in Dresden Abh., 

 1894, p. 12, pi. 1, fig. 41. 



' Idem, p. S, jil, I, figs. 3G-.3S. 



"Engelhardt, Ilermann, Senckenbergischc naturf. GesoU. Abh., vol. 

 19, p. 19, pi. 2, figs. 15 and 16, 1895. 



Oceurrence. — HoUy Springs sand. Early 

 Grove and HoUy Sjirings, Marshall County, 

 Miss, (collected by E. W. Berry). Lagrange 

 formation (in b(^ds oi Wilcox ago), Puryear, 

 Henry County, and Pinson, Madison County, 

 Tenn". (coUected by E. W. Berry). 



Collections. — U. S. National iluseum. 



Genus C/ESALPINITES Saporta. 



C^SALPixiTEs PiNSONENSis Berry, n. sp. 



Plate L, figure 13. 



Description. — Leaflets smaU and sessile, at- 

 tached somewhat obliciuely, broadly elliptical 

 in outline, about 7 millimeters in length by 4.5 

 millimeters in maximum width in the basal 

 half. Apex broadly rounded, with a mucro- 

 nate point at the end of the midrib. Base 

 l)roadly rounded, somewhat ineciuilateral, and 

 broader than the apex. Margins entire. Tex- 

 ture coriaceous. Venation immersed, even the 

 midrib scarcely discernible. 



This small, almost orbicular leaflet is clearly 

 allied to Csesalpinia. It is sparingly repre- 

 sented at Pinson and the material collected 

 shows only the upper surface of the leaflets, so 

 that the venation characters can not bo made 

 out. It is much smaller than most of the forms 

 of Ca'salpiniaccse and Mimosacese described 

 from tlm Wilcox deposits and is not close to any 

 previously described forms. It suggests some- 

 what Cfesalpinia sellardsi Berry, a true Cjesal- 

 pinia, which comes from the Alum Bluff forma- 

 tion of Florida, and in which the leaflets were 

 more inequilateral at the base and conseciuently 

 borne at a more oblicjue angle on the rachis. 



A nuiuber of species described from later 

 Tertiary liorizons of Europe arc similar to the 

 form under discussion, for example, the lower 

 Oligocene forms (Stampian) of Csesalpinia 

 toivnsliendi Heer. 



Cxsalpinites pinsonensis comes from the 

 basal sands of Holly Springs or middle Wilcox 

 age near the eastern boundary of Madison 

 County, Tenn., and is of especial interest on 

 that account. It is very close but somewhat 

 larger than a form from the Tertiary of Bolivia 

 described by Engelhartlt ' as Desmodium cllip- 

 ticurn. 



' Engelhardt, Hermann, Naturwiss. Gesell. Isis in Dresden Abh., 

 1894, p. 8, pi. 1, figs. 42-44. 



