KNowi.TON] DESCRIPTION OF SPECIES. 57 



Bekberis^ (;i(!Antp:a n. sp. 



m. XI, Hg. 1. 



Leaf of larcve size, very thick and leathery in texture, pahnately 

 deeply three-lo})ed, lobe.s lanceolate, provided with numerous, large, 

 trianj^ular or trian<j;"ular-lanceolate, sharp-pointed lobes; lobes pro- 

 vided with very thick, fleshy midrib, and an apparently contimious 

 intramarg'inal vein; secondary branches very thin, arising- from the 

 midrib and passing directly or with a slight curve to the apices of 

 the lateral lobes; liner nervation very thin, anastomosing, producing 

 large, irregular areas. 



The single magnificent specimen upon which this opinion is based 

 remains absolutely unique. It was clearly a thick, leathery leaf, and 

 pahnately deeply three-lobed. There is no means of knowing the full 

 length and width, as both base and apex are absent. Of the middle 

 lobe only about cm. is preserved; of the lateral lobes 8 to 11 cm. is 

 preserved. The widest part between the lobes as now preserved 

 measures about 13 cm. When perfect it was probably at least 15 

 cm. long and spread probably 18 cm. The petiole is, of course, not 

 preserved. The nervation, as set forth in the diagnosis, is strongly 

 marked. There is a thin intramarginal vein running around the entire 

 margin, this l)eing especiall}' prominent in the interval between the 

 lateral lobes. Each lobe is provided with a very thick, fleshy midrib, 

 from which arise, at irregular intervals, the thin secondaries which 

 pass to the sharp points of the lateral lobes. The other nervation 

 produces large anastomosing areas. 



I am somewhat luicertain as to the proper generic reference for this 

 fine leaf. The late Dr. Newberr}^, to whom the specimen was sent 

 and who had prepared a drawing of it, placed it provisionall}^ in the 

 genus Cnicus, or Carduus, as it is now called. The peculiar sharp 

 lobes, as well as the flesh}^ midribs, are suggestive of the large spiny 

 leaves of thistles, but no species of this genus, so far as I know, has 

 palmate leaves. The individual lobes resemble the whole leaf of many 

 species of Carduus, but the palmate character efi'ectively excludes it. 



The only genus with which I have been able to satisfactorih^ compare 

 this leaf is Berberis, and the nearest species is B. trlfollolata Moric. 

 This species, now found in southern Texas and Mexico, has ordinarily 

 small trifoliolate leaves, but occasionally one is found in which two 

 and sometimes all three leaflets are united for a greater or less distance 

 at the base, thus producing a leaf of exactl}' the same character as the 

 fossil under discussion. The individual leaflets of this living species 

 are rarely more than 4 cm. in length, whereas the lobes of this fossil 

 form must have been 10 or 12 cm. long. In the matter of lobation, 

 nervation, and general appearance, even to the intramarginal vein, 

 the agreement l^etween them is perfect. For this reason 1 have placed 

 this fossil under Berberis and denoted the doubt by a question mark. 



