(89) 



Lesquereux separates the Dakota leaves of Sterculia from 

 Aralia merely on account of the " primary divisions and pri- 

 mary nerves from the top of the petiole." This character, 

 which I do not consider diagnostic, would refer this leaf to 

 Aralia as the lateral primaries branch from the midrib a con- 

 siderable distance above its base. The venation is some- 

 what similar to the Dakota Sterculia reticulata Lesq. (Fl. 

 Dak. Group, pi. 34. f. 10), and also to that of Aralia trans- 

 z'crsincrvia Sap. & Mar. described by Hollick from Oak- 

 neck, Long Island (Bull. Torrey Club, 21 . 54. pi. 176./. i . 

 1894) which leaf he does not consider an Aralia. 



In outline this leaf resembles Sterculia lugubris Lesq. ex- 

 cept that the primaries are not basal ; whether the lobes were 

 produced to the length they are in that species is of course 

 conjectural. Our fragment is also somewhat similar in out- 

 line to the fragment (Fl. Amboy Clays, ^/. 26./. 2) referred 

 by Newberry to Aralia quinquepartita Lesq., in which how- 

 ever the venation is unfortunately obliterated. Hollick's 

 Sterculia sp. (/. c.) probably belongs here ; his f. 4 is the 

 fragment of a much smaller leaf, but his larger fragments 

 (y. 5-7) might well be the acutely tipped lobes of our leaf, 

 the venation of the two corresponding very well. 



Our leaf also has somewhat the appearance of Aralia 

 Jorgenseni Heer (Fl. Foss. Arct. 7: pi. loi. f. /) but the 

 sinuses are not quite so deep. It might further be compared 

 to Aralia Wellingtoniana Vaughanii Knowlton from the 

 Woodbine formation of Texas (Dakota). The latter is tri- 

 lobed, the lobes slender and entire ; not figured, however. 

 (Knowlton; Hill, Ann. Rep. U. S. Geol. Surv. 21: 318. 

 1901.) 



Sterculia Snowii bilobata var. nov, PI. 4.3./. 7. 



Sterculia Snowii \s known from the Dakota Group of Kan- 

 sas and New Mexico and the Cheyenne Sandstone at Belvi- 

 dere, Kansas. The specimen from the Matawan here figured, 

 in its outline approximates Liriophylhim Beckzt'ithii Lesq., 

 from the Dakota Group, but the venation is radically differ- 



