GREENE PTELEA IN THE WEST AND SOUTHWEST. 51 



their twigs of one 3Tar\s growth; and of actually supremo taxonomic 

 importance in the case of the "oaks arc the color, texture, duration, 

 marginal indentation, and pubescence of the leaves. Not even the 

 characteristics of the acorns are found to he of equal weight with the 

 mere hue and texture of foliage in the classification of the oaks. And 

 all this I find true in regard to Ptelea, and even more; for the charac- 

 teristics of twigs of one season's growth in tliis genus, their colors, 

 kinds, and degrees of pubescence, evenness and unevenness of sur- 

 face, etc-, are many times more diverse than thc}^ are in any oaks; 

 and both those sets of characters — those furnished by the twigs and 

 those presented by color and texture of foliage— either set indispen- 

 sable to any natural arrangement of Ptclea species, are liorc for the 

 first time brought to notice. The chaos that has reigned hitherto in 

 respect to Ptelea of the farther West and Soutliwest luis held sway 

 because it has not been seen tiiat, in the species of one region, the 

 twigs are chestnut-brown and velvety in one set, chcstnut-bruwn and 

 smooth and shining in another set; while in another and remoter dis- 

 trict all the species have cinnamon-red wai'ty twigs; and in a third 

 group the twigs in all the species are either yellowish or straw-colored 

 or nearl}^ white and in almost all smooth and shining. I say with con- 

 fidence that these marked diversities which even the dead and dry 

 herbarium specimens exhibit can not have been looked at; for no 

 botanist would pretend that one species of shrul) or tree could so vary 

 in respect to the bark of its twigs and branches. 



The fruits in this genus are also found to present a considerable 

 array of characters available for specific diagnosis, and also even for 

 the grouping of the species; and some new descriptive terms have 

 seemed to be called for in connection with them. The body of the 

 samara, while in the broad, thin-leaved species it is thin and rather 

 flat as w^cll as small in proportion to the wing, is by comparison large 

 and double-convex as w^dl as more narrowly winged in the species 

 that have a thick and subcoriaceous foliage. This seed bearing body 

 is in some marked by ruther closelv parallel transverse rid<rGS. with 

 lines of gland dots running between them, or else the ridges are irreg- 



ularl)^ broken and run into a reticulation, with one or more dots in the 

 middle of each mesh. In either instance the ridges may, at the edge 

 of the body or a little beyond it, unite to form a wall more or less 

 definitely surrounding the body — which wall 1 denominate the circum- 

 vailation— or they may x^ass directly into the reticulation of the wing 

 itself, leaving the body without circumvallation. 



In the Californian group of species the ridges of tlie body of the 

 fruit are mostly faint or obsolete, in wdiich case the gland dots arc 

 multiplied and very conspicuous, in the Lower Californian nearly or 

 quite wingless species rising into a prominent tubi^rcidation. Again, 

 and with respect to its proximity to cither the }>ase or the suuunit of 



32966—06 3 



