CrCADACFJE. 21 



usually been overlooked as a means of identification. The in- 

 clination of pinnae to the rachis, and their alternate or opposite 

 disposition are characters which have been used as the basis of 

 specific determination, but such features as these are likely to prove 

 misleading unless used with great caution. In one part of a frond 

 the pinnae may be distinctly opposite, and in another alternate. 

 The same kind of variation in the angle of insertion of a segment 

 to the rachis, is readily seen in the large fronds of such recent 

 species as Ceratozamia mexicana, Brong., Macrozamia Maclcayi, Miq., 

 and many others ; also among fossils in the larger specimens of 

 Zamites Buchianus (Ett.), etc. A comparison of the young and 

 old fronds of many cycads reveals the same striking difference as 

 regards the inclination of the pinnae. The open or closely set 

 arrangement of pinnae is another misleading character ; e.g. in an 

 old frond of JEnceplialartos longifolius, Lehm., the pinnae are for 

 the most part in contact with one another, but the j'oung frond 

 presents a distinctly open habit, with the pinnae much more 

 openly arranged. In Oto&amites Goppertiamts (Dunk.) there is 

 the same difference in this respect between the upper and lower 

 portions of the same specimen, e.g. PL I. Figs. 1 and 2. 



The form of the epidermal cells is a character of doubtful 

 value, and at the same time one which can only be made use of 

 under favourable conditions of fossilization. The custom of asso- 

 ciating spiny margins with the pinnae of Encephalartos has led an 

 American writer to adopt this feature as the leading characteristic 

 of his genus Encephalartopsis. 1 Fontaine has founded this new 

 genus on some very fragmentaiy and imperfect pinna? with spinous 

 margins and anastomosing veins. None of the figured fragments 

 afford any clue as to the nature of the pinna base, or as to the 

 manner of insertion on the rachis. The material is hopelessly 

 inadequate for the institution of a new genus. The fact of the 

 fragments possessing anastomosing veins deters Fontaine from 

 including them in the recent genus Encephalartos ; as it is, he 

 prefers to institute a new term, and to consider the species as 

 probably a "prototype" of the recent genus. It is true one is 

 accustomed to associate spiny pinnae with species of Encephalartos, 

 but there are several forms of that genus in which no indication 

 of such a character is found ; and on the other hand, spiny pinnae 



1 Fontaine (A. 2), Potomac Flora, p. 174. 



