Maryland Geological Survey 489 



present connection. One feature should be emphasized, however, and that 

 is the variability in lobing which these leaves show. They are primarily 

 palmately trilobate in character, but show a considerable range in varia- 

 bility in the development of subordinate lobes on any or all of the princi- 

 pal lobes. The point of insertion of the secondaries and the character 

 of the apices, whether obtusely rounded or acutely pointed, are also 

 very variable characters. Their coriaceous texture should also be 

 referred to. ■ 



This variability in lobation in AralicepJiyllum finds an analogy in the 

 like variation among the leaves of the modern Sassafras as pointed out 

 by the writer in 1902.^ Sassafras is predominantly entire or trilobate 

 but individual leaves can be found showing every stage of incipient loba- 

 tion up to leaves with five or even six lobes which are strikingly similar 

 to the leaves of AraUcephyllum (compare the plates of the latter with loc. 

 cit. figs. 2-4) . 



In the original treatment of the genus four species were recognized. 

 These have here been reduced to two species with which have been united 

 certain forms described by Fontaine as Aceriphyllum, Platanophyllum, 

 and Hedercephyllum. The material is all rather fragmentary and is 

 strictly confined to the Patapsco formation. 



With regard to the botanical affinity of these forms, like so many of 

 the Lower Cretaceous dicotyledones, little can be said of a decisive 

 nature. They are comprehensive types and unite the foliar characters 

 of leaves which in the Upper Cretaceous are referred to Aralia, Sassafras, 

 Platanus, and Cissites. As the genus Aralia is used by paleobotanists 

 it would naturally include the following forms, although there is a 

 distinct advantage in the present case in keeping them separate as con- 

 stituting the genus AralicepJiyllum which indicates without undue defi- 

 niteness their supposed affinity with the modern Araliacese. 



One of the species, AraUcephyllum crassinerve (Fontaine), shows 

 considerable resemblance to the Earitan form Aralia Newierryi Berry 

 and was doubtless ancestral to it. 



^ Berry, Bot. Gazette, vol. xxxiv, 1902, pp. 426-450. 



