43^ BOTANICAL GAZETTE [december 



say. It might serve to strengthen the leaf and prevent tearing 

 where the lobes are large; at any rate it is an acquired feature 

 the origin of which is shown in some specimens and serves to 

 prove that the ancestral leaf was simple, and not lobed as Ward 

 holds; it also rules out the contention that ancient leaves lack- 

 ing this arrangement are not forms of Sassafras, as w^e would 

 hardly expect this character in the ancestral forms becoming 

 lobed. 



This marginal vein at the bottom of the sinus was lacking in 

 fifteen single instances, and a secondary sinus lacked it in four 

 instances out of a total of 201 leaves examined especially because 

 of their diverse lobation. In nature it is undoubtedly a nearly 

 constant character. In some cases the secondary which runs to 

 the sinus is continued into the upper or the lateral lobe, and a 

 tertiary from the midrib unites with one from the primary, then 

 running to the sinus in the usual manner; or it may be a branch 

 from the secondary together with one from the primary which 

 forms the margin; the latter, however, is always thickened irre- 

 spective of the system to which it seems to belong. I have col- 

 lected several specimens in w^hich the secondary forks before 

 reaching the sinus, the branches striding it after the manner 

 which usually obtains in the lobed leaves of a variety of genera, 

 the branches connecting with the branches from the next sec- 

 ondary above and with those from the primary below. 



No fossil American Sassafras has, orat least shows, this pecul- 

 iar marginal venation, but it occurs in a European representa- 

 tive of this genus according to Ward ?. The secondary seems 

 to run to the sinus in S. bilobatiun, S, platanoides, and in some 

 specimens of S, progenitor. The form which Lesquereux describes 

 as Aralia acerifolia^ approximates this style of venation as does 

 also the leaf referred to Lindera ve?msta^, which leaf Knowlton 

 refers to Benzoin. 



On general principles we would expect the ancestral Sassa- 

 fras to have had simple, lanceolate, or ovate-lanceolate leaves 

 with entire margins. Not only is this a rule among the Laura- 



^ Sassafras Ferrettianum Mass., from Senegal. Fl. Foss. de Seneg. //. 12. fig* /• 



8 



Cret. and Tert. Fl. pL ^9. fig, j. 9 See Fl. Dak. Group, pL lb. figs. 1-2 



M 



