440 BOTANICAL GAZETTE Ldecember 



narrower and more produced lobes ; the secondaries, while indis- 

 tinct, seem to conform to those of Lesquereux's other specimens, 

 thereby differing from Newberry's leaves. The two series are, 

 furthermore, widely separated geographically and are from 

 different geological horizons. As previously stated, the Amboy 

 clay leaves show no affinity to Sassafras; especially is this so in 

 the very small, probably young leaves. Whether the Dakota 

 group forms are Sassafras or not is not easy to decide. No 

 modern Sassafras leaves which I have seen have the primaries 

 and the lateral lobes so nearly horizontal; the secondaries are 

 not so uniformly regular, nor do they curve upward to join the 

 next above at a point. In the modern leaf an outwardly and 

 downwardly directed branch from the latter is emphasized. 

 There is never such an open sinus, amounting as it does to nearly 

 90°, and the lobes in the modern leaf have their margins inflated 

 and not straight. In these ancient leaves the sinus never has a 

 marginal vein, the secondary in this region forking and striding 

 it, or curving to join its neighbor. In all cases the secondary 

 system seems to be uniform throughout the leaf, while in the 

 modern leaf there is always evidence of changed conditions in 

 that region around the sinus ; the secondaries or their representa- 

 tives from both the primaries and midrib are changed in size and 

 direction and usually belong to the tertiar^^ system. None of the 

 Dakota leaves of this species show the characteristic basal vena- 

 tion of the modern leaf. While we should not, necessarily, 

 expect Cretaceous species to conform to the modern type, still 

 the character of the secondary system in the former is so different 

 from what would obtain in a leaf descended from a simple 

 ancestor, such as we hold Sassafras to have done, that we are 

 inclined to associate these leaves with those trilobed forms which 

 have been referred to Aralia, laying aside, for the present, any 

 consideration as to whether or no they are true species of Aralia. 

 Lesquereux originally considered his leaves to be a variety of 



Miidgi 



Leaves 



which have been referred to this species (5. aciitilobtmi) have been 

 found in both the Lower and Upper Cretaceous, and at the fol- 

 lowing localities: Woodbridge and Cliffwood, N. J.; near Ft. 



