IQ22.J F. F. Laidlaw : Indian Dyagonfiies. 387 



Clear-cut venational characters are not to be found, though 

 this is possibly to some extent due to the fact that the very large 

 genus Gomphus is something of a " dump" for a certain number 

 of species whose exact position is at present doubtful. On the 

 other hand, anyone who examines a considerable number of 

 specimens belonging to several of the genera of the series can 

 scarcely fail to notice that whereas the specialization of cross- 

 veins between M,_j and Mj,, used by Williamson as the most 

 important character for the definition of the series, is tolerably 

 constant in the hinder-wing, in the case of the fore-wing a certain 

 range of specific and even of individual variation may occur. 

 Further, I believe that in certain genera, probably in those which 

 are to be regarded as the most highly evolved of the series, this 

 variability in the case of the fore- wing is by no means marked ; 

 it seems to me even to be absent in some cases, so that in such 

 genera the specialization of the fore- and hinder- wings is about 

 equal. Unfortunately, owing to the occurrence of tolerably nu- 

 merous cases of individual variations of the fore-wing, the degree 

 of specialization exhibited by it cannot as yet be used as a generic 

 character. 



Of characters available for grouping genera which appear to 

 form natural assemblies in the series the most evident are those 

 supplied by the anal appendages of the males. These characters, 

 backed to some extent by colour peculiarities, and also here and 

 there by features of the venation, give a grouping which I am per- 

 suaded is tolerably natural, and of some practical value. I would 

 again insist that in presenting such an arrangement here I am 

 drawing largely on suggestions and hints made by de Selys, by 

 Williamson and by Ris, and that I am dealing entirely with Indian 

 or Oriental forms. I believe the tribes or sections of the series 

 defined below to have approximate^ equal value. Davidius 

 would seem venationally to be the least specialized, and perhaps 

 the most easily defined. Onychogomphus and Heterogomphus are 

 the most advanced. In them it is rare or exceptional to find an 

 individual in which the specialization of the cross- veins between 

 Mi_ 3 and M, of the fore-wing is not as fully developed as in the 

 hinder-wing. 



I propose then to group the bulk of the oriental members of 

 the series in the following five " groups," leaving out of account one 

 or two genera whose position is to my mind doubtful, or which are 

 not sufficiently known to me. From what I have seen of the 

 British Museum collection I should be inclined to say that there are 

 still several genera awaiting recognition, especially amongst the 

 material from Tonkin, a country apparently very rich in Gomphinae. 

 Groups : ' 



Davidius, 



Cyclogomphus, 



Gomphus, 



Onychogomphus, 



Heterogomphus. 



