43^ BOTANICAL GAZETTE [December 



they are carried upward, while new laterals are added toward 

 the tip of the blade. It may be that the diagonal position of 

 the leaves in the bud causes pressure at that particular lateral 

 part near the tip of the primary, causing more or less atrophy of 

 that part of the blade. This tendency once inaugurated the rest 

 is simple, for those portions of the leaf at the tips of the prima* 

 ries would have nearly all their leaf-forming energy expended in 

 increasing the length of the lobe ; possibly especially good envi- 

 ronment was a factor in the original lobation, as witness the great 

 development of the lateral portions of the leaf blade in the five- 

 lobed forms occurring in rich soil. The Potomac species under 

 discussion bears some resemblance to certain species referred to 

 Sterculia, as well as to the asymmetrical terminal leaflets of 

 some compound leaves, but I feel that we are justified in con- 

 sidering it a true Sassafras — the first that we can identify as 

 such with any degree of certainty. As pointed out by Fontaine 

 in the fragment of this leaf figured (by him), the opposite pri- 

 mary is considerably stouter than its fellow which runs to the 

 lobe which is preserved, lending color to the supposition that 

 this species was also trilobed. 



The next forms which we have to consider are those recorded 

 from the Cheyenne sandstone of Kansas, They were probably 

 contemporaneous with the foregoing or possibly somewhat more 

 recent, and flourished about looo miles to the westward. 



Sassafras sp. Knowlton in Hill, Am, Jour. Sci. i : 212. 

 1895, represents fragmentary remains of doubtful identity from 

 Belvidere, Kansas. 



Sassafras obtusum Lesq., fi/st described by him as Popiihtes 

 salhburiaefolia in 1868, then as Sassafras, and then as Cissites, 

 and Sassafras cretaceum obtusum Lesq. had both better be 

 referred to Cissites ; at least this is a more natural place for 

 them among the Lauraceae. Newberry includes them both in 

 his composite species 5. cretaceum. They resemble somewhat 

 the trilobate forms of Cissites harkerianiis Lesq., and also *C. 

 insignis Heer, and there is a distant resemblance to Dawson's 

 Sassafras Selwynii from the Canadian Upper Laramie. They are 

 both found in the Dakota formation, and S. cretaceum obtusion 



