16 CRETACEOUS AND TERTIARY FLORA. 



which, in some specimens larger than the one figured, are now deeply cut 

 by divisions like pointed lobes, there is still another character, remarked 

 on specimens recently discovered, which seems more forcibly to separate 

 these forms from the Laurinece, and indicates a more evident relation to 

 the Araliacece. A number of those specimens communicated by M. Chs. 

 Sternberg, to whose careful and zealous researches the Flora of the Dakota 

 Group is indebted for many important discoveries, represent large leaves, 

 which, by the outlines, the nervation, and the dentate borders of the lobes, 

 are like S. mirabile of pi. xii, fig. 1. The leaves, however, which are much 

 larger, the lobes measuring as much as ten centimeters in length from 

 the point of union of the primary nerves, greatly differ by the forking 

 of the lateral nerves from a point two and one-half centimeters above 

 their base, thus forming, of course, a subdivision of these lobes into 

 two equal parts, or a palmately five-lobed leaf. They are described as 

 Sassafras (Araliopsis) dissectum. Among the innumerable varieties in the 

 shape of the leaves of the living Sassafras officinale we see a constant and 

 gradual mode of division, passing from a round or oval and entire shape 

 to a bilobed and trilobed one; but, as yet, I have been unable to observe 

 a single case of subdivision of the lateral lobes, or to find a palmately 

 five-lobed Sassafras leaf. This character is, on the contrary, far more 

 generally seen in the Araliacece of our time. Even in a section of the 

 Araliacece, the genus Hedera, whose leaves may be compared to some of 

 those under examination, I do not know any species with trilobate leaves. 

 Hedera turbascens, H. discolor, H. argentea, H. aurifolia, H. jatropcefolia, 

 have leaves five to seven palmately lobed, or when occasionally trifid their 

 segments are narrow and acuminate. From this the relation of the five 

 palmate leaves to the Araliacece becomes more evident. 



Going further into this kind of investigation, we are met by a new 

 difficulty in the appearance of another modification in the character of 

 this peculiar type of leaves. In examining the first specimens of the 

 species represented (pi. xii and xiv), I could but consider them as repre- 

 senting either Sassafras (Araliopsis) obtusum or S. mirabile, the specimens 

 being fragmentary, having only the lobes or part of them preserved. As 

 long as the auricled and peltate base was unknown, the reference of the 

 specimens could not be different. The nervation, the form of the lobes, 



