Cretaceous species Liriodendron semialatum Lesq.* and while 

 Fig. 13 at first sij^ht suggests Aralia, Cissites or some other but 

 little understood fossil leaf, it would be the logical successor of 

 the semialatum form, being a more robust leaf with a shortened 

 length and an increased breadth. It is however a remarkable 

 leaf to have been borne on a tulip-tree and was sent to me from 

 Columbus, Ohio by Mrs. W A. Kellerman, an amateur botanist 

 of that place. 



It has become more and more evident to paleobotanists that 

 many of the numerous leaves variously referred to Credncria, 

 Cissites, Araliopsis, Grcwiopsis, Sassafras, Plata m/s, etc., are, at 

 least some of them, unnaturally identified and their true affinities 

 but little understood ; and while perhaps all of these and other 

 genera are badly in need of revision, it would be rash to attempt 

 one without far more material than is at present available. In 

 this connection several of our leaf specimens of Liriodendon 

 Tulipifera are particularly interesting ; at the same time I do not 

 feel justified in anything further than calling attention to them. 

 The first, Fig. 12, shows a very anomalous leaf, one which al- 

 most exactly corresponds with the Cissites aciiminatus of Les- 

 quereux.f It stands alone in its uniqueness, and yet the tree 

 which furnished it bore many leaves of a similar general shape 

 and with similar venation ; they were of all sizes, some of the 

 specimens being 130 mm. in length, and all were like the speci- 

 men in question except that the acute apex was cuneate or with 

 a wide obtuse sinus, the resulting apical portions of the blade 

 showing a slight tendency to become lobed. 



Another specimen which is of interest in this connection is 

 Fig. 16, on which I will offer no comment other than to call 

 attention to its resemblance to Cissites obtusilobtLS Lesq.| From 

 this leaf I have a complete series showing a gradual shortening 

 of the midrib and a gradual lengthening of the lateral lobes, ending 

 in the curious form shown at Fig. 18 in which the leaf consists of 

 a single orbicular lobe on each side, the blade being nearly four 

 times as wide as the midrib is long. Fig. 10 shows a leaf which, 



* Fl. Dak. Gr. 204. //. 2j, f. 2-4 ; pi. 2q, f j. ( 1891 ) 1892. 



| Compare with fig. 3 on pi. 5, Cret. & Tert. Fl. 1883. 



+ Compare with fig. 5 on pi. 33, Fl. of Dak. Group. (1891) 1892. 



