4 6 GLOSSOPTERIS. 



to believe belonged to the same plant. Some degree of variation 

 in these characters has been admitted by all. Bunbury, 1 one of 

 the earliest writers on the Indian Glossopterids, has called attention 

 to the fact, and in more recent times Professor Zeiller and 

 Mr. Seward have given further illustrations. 



Indeed, a detailed examination shows that there are hardly any 

 characters which may be regarded as constant in fronds of the 

 same species. Size and shape are admittedly dangerous guides, 

 and it would seem almost certain that some of the narrower fronds 

 may have been borne on the same rhizome as those of broader 

 dimensions. 



The characters of the midrib, its thickness, its persistence or 

 non-persistence at the apex, and the obtuse or acute termination 

 of the leaf are often simple corollaries of the size and shape of 

 the frond. The angle of divergence of the secondary veins from 

 the midrib is quite untrustworthy, and frequently varies in different 

 parts of the same frond, and on opposite sides of the midrib 

 (Text-figs. 16, 17, and 19). The one character which does appear 

 to be fairly constant, though it is by no means without a certain 

 amount of variation, is the average openness or closeness of the 

 secondary nerves, and consequently the shape of the meshes or 

 areoles. 



In attempting a general revision of this large genus, the choice 

 lies between grouping broadly, and distinguishing in detail. "What- 

 ever classification may be adopted must necessarily be an artificial 

 one, until we know more of the fructification of the different 

 types of frond. Since an artificial classification is without 

 alternative, and seeing that the value of such a classification 

 depends largely on its convenience, it would seem more convenient, 

 as well as more in accordance with our knowledge of the variations 

 in the details of the fronds of this genus, to maintain com- 

 paratively few species, and to group together those fronds which, 

 though they may differ in one or more details from the typical 

 form, are not sufficiently dissimilar in the aggregate of their 

 characters to warrant separate specific rank. Some authors have 

 expressed this view by distinguishing a number of varieties or 



1 Bunbury (Gl), p. 327. 



