MARYLAND GEOLOGICAL SURVEY 877 



number, separated by relatively narrow ultimately rounded sinuses, com- 

 prising an o\ 7 ate medium terminal lobe and two main lateral lobes on 

 either side, the lower pair being more or less divided. In the Maryland 

 material the auxiliary lobe on the lower side of each main lateral lobe is 

 feebly developed. In the Greenland material it is at least half as large 

 as the main lobe and the separating sinus extends half-way to the base. 

 Petiole stout, its full length unknown. Midrib very stout and prominent, 

 straight. Lateral primaries two on each side, stout and prominent, the 

 lower pair subopposite and suprabasilar, the upper pair sometimes sub- 

 opposite, oftener separated by a wide interval. The lower primary may 

 fork a short distance above its base, as it does in the Greenland material 

 after an interval of only about 1 cm., or this fork may be at least 4 cm. 

 above the base as in the Maryland material, the distance depending on the 

 extent to which the auxiliary lobe is developed. The angle of divergence 

 of the primaries from the midrib is about 40, but varies from specimen 

 to specimen, the basal pair in general being somewhat more divergent than 

 the upper pair. The secondary and tertiary venation is usually obsolete. 

 Some specimens show a few thin remote secondaries diverging from the 

 primaries at angles of about 45 and sweeping upward in ascending camp- 

 todrome curves. 



This species was described by Heer from the Greenland Upper Creta- 

 ceous (Atane beds) and has been found by the writer in the Magothy for- 

 mation of both New Jersey and Maryland. The fragments from Marthas 

 Vineyard, Massachusetts, and Tottenville, New York, identified as this 

 species by Hollick, 1 are not this species in the writer's judgment. There is 

 a great display of Aralia-like forms in the Middle Cretaceous both of this 

 country and Europe, and these forms are especially abundant in the 

 Dakota sandstone of the West. Among this diverse display of forms 

 Aralia ravniana is approached in both size and outline by Aralia towneri 

 Lesquereux of the Dakota sandstone, which is to be regarded as a closely 

 related geographical mutant. Comparisons with existing plants are not 

 so satisfactory, although many tropical Araliacece show suggestive resem- 



1 Hollick, Mon. U. S. Geol. Survey, vol. 1, p. 99, pi. xxxvii, figs. 1, 2, 1907. 



