DESCRIPTION OF SPECIES— OELTIDE^. 191 



Habitat. — Castello's Ranch, Colorado {Capt. Bcrthoud, Dr. F. V. Hay- 

 den'). Very rare. 



OELTIDE^. 



CELTIS, Tournf. 



Considering its American representatives, this genus is distinctly charac- 

 terized by petioled leaves, short, truncate or subcordate at the more or less 

 inequilateral base, ovate or oblong-lanceolate, acuminate, with serrate, creuate, 

 or entire borders, triple-nerved. The two lateral basilar nerves ascending 

 to the middle of the leaves or above, are camptodrome, like their division 

 and the secondary nerves also. 



More than seventy species of this genus are described in De Candolle's 

 Prodromus as distributed over the whole world. The number, however, is 

 exaggerated; for the forms are disposed to vary greatly by culture and to be 

 modified by various atmospheric changes. For example, twelve species are 

 credited to North America, and these, from the opinions of botanists of the 

 present time, who have had opportunity to compare them and to follow their 

 variations, are reduced to two species, one, C. occidentalis, which ranges in 

 longitude from the Atlantic to the base of the Rocky Mountains, and in 

 latitude from New England to Florida, and the other, C. pallida, Torr., which, 

 inhabits Texas and Mexico. The first of these species is locally very abun- 

 dant. 



The geological records of Cellis are for this continent as yet obscure. . 

 The flora of the Mississippi Eocene has two fragments of leaves, described 

 as Celtis breviforta,Les<:\\. (Trans. Am. Philos. Soc, vol. xiii, p. 41 6, pi. xx, figs. 4 

 and 5). The lower part of these leaves and their nervation show the characters 

 of Celtis; but the upper part is destroyed, and thus the generic relation is not 

 positively ascertained. In Europe, which has one living species only in its 

 flora, paleontology has recognized seven species of Celtis, all Miocene. 



MORE^. 



FICirS, Tournf. 



The leaves of species oi' Ficiin are so variable in size, shape, consistence, 

 nervation, etc., that it is not possible to expose any characters by which their 

 reference to the genus may be positively ascertained. Schimper remarks on 

 this subject (Pal. Veget., ii, p. 728): — "That most of our determinations of 



