46 BOTANICAL GAZETTE I JULY 



alone are always more or less problematical, and yet in many 

 cases form alone has served for specific distinction. 



Venation also is much depended upon in defining species, 

 although both venation and leaf f^rm are characters which vary 

 in a marked degree in the modern tree, even on the same indi- 

 vidual. Leaving out the anomalous forms from shoots, forced 

 buds, saplings, etc^ almost any sassafras or tulip tree will show 



eV^ariati-^n amctv^ mature leaves on branches 



tj^q^ret 



branches) should l)ear typical 

 eries from different trees or of 



different ages the variation in shape and minor characters or 



venation is almost endless. Perhaps no style of venation is 

 more characteristic generically than the peculiar form that 

 obtains in Liriodendron. That it cannot be relied upon for spe- 

 cific distinctions we know from its wide variation on the livmg 



i 



tree, as well as among the fossil species referred to this 

 genus. In some specimens the secondaries are opposite 

 instead of irregularly alternate ; in some they are parallel 

 almost to the margin and ascending ; while in others they soon 

 divide. The angles of divergence of the secondaries show 

 every degree of divergence, acute angles, right angles, or obtuse 

 angles. All that should be claimed for venation characters, 

 generally speaking, is usefulness in generic distinction or as sup- 

 plementary evidence in connection with species. 



Holm considers nearly all of the described forms of Lirioden- 

 dron invalid, while most authorities consider nearly all of them 

 perfectly good species. Professor Ward's dictum, that for geo- 

 logical purposes it is not so much a question of correct botanical 

 determination as the correct recognition of a plant once named and 

 associated with a given deposit, is quite true ; but it loses its force 

 when we are considering forms with variable leaves, unless each 

 variety is constant and peculiar to a given formation, which is 

 obviously not the case. In pursuing the comparison through the 

 intermediate forms of ancient Liriodendrons, who can say where 

 to break the chain of gradations for the boundary of separate spe- 

 cies; among the gradual modifications of form,strikingly similar to 



