4 PROCEEDINGS OF THE NATIONAL MUSEUM VOL. CG 



in size and prominence proximad, until at the base Avbere it joh\s 

 the petiole it is 3 millimeters in diameter. The .secondaries sue ilvn 

 but well marked, closely spaced and ascendinii'. Tliey diverge fro!n 

 the midrib at angles of about 30 degrees, are relatively strnidit iji 

 their courses, although adjacent ones occasionally join on thoii- way 

 toward the margin, where their extremities are united ])\ a looped 

 marginal vein about 1 millimeter within the margin. 



There cannot be the slightest doubt but that these leaves represent 

 the genus Clusiu^ presenting as they do to the last detail the foliar 

 characters of this genus. I have comi)ared them Avith the leaves of 

 all of the existing forms from equatorial America Avhich are repre- 

 sented in the National Herbarium. There are 3 existing species in 

 the Trindad flora whose leaves are extremely difficult to distinguish 

 from the fossil. These are Glusia martini Sagot and Glusia palnii- 

 cida L. C. Rich which are large trees of the forest (specimens from 

 Balandra Bay), and tlie wide ranging C'lfHta rosica Linnaeus, a 

 somewhat smaller tree of the rocky coasts, at least the specimens 

 from Trindad are so labeled. 



The foliar characters of these are very convergent. In general 

 the leaves of Cusia palmicida are somewhat more elongated. I 

 doubt if it is possible to certainly distinguish Clvsla rosea and 

 Clima martini from the leaves alone. The fossil appears to be more 

 nearly identical with latter than the former of these, but occasional 

 leaves of the former are indistinguishable. This being the case the 

 fossil form is described as Clusia fosailia, and it is sugge^sted that 

 it might very well represent the stock subsequently differentiated 

 into the three living species mentioned above as recognized by 

 modern systematists. 



So far as I know this is the first fossil species of this interesting 

 genus to be recognized, although I have a fine and much larger 

 species from tlie later Miocene of Trinidad. 



noJofj//H'.~-CAiL Xo. 37010, V. S. N. M. 



Order THYMELEALES 



Family LAURACEAE 



Genus PERSEA Gaetner fils 



PERSEA AMERICANA Miller 



Plate 1, lis. 4 



The single incomplete leaf figured is the only specimen of this 

 species collected. It has the form and venation of various tropical 

 species of Phoehc and Persea and appears to be identical with the 

 leaves of the existing Persea americanr/ Millei'. 



