1922.] F. F. Laith.aw : Indian Dragon/lies. 399 



mark at upper end of first lateral suture, very narrow complete 

 black line on second suture. Abdomen with segments 1-6 black 

 with longitudinal brown bands dorsally ; 7-10 mainly brown. 

 Hindermost femora yellow. 



Length of abdomen 38+ 1*25 mm., of hinder-wing 30 mm. 



The species P. fcae Selys, from Burma, is apparently closely 

 allied. It differs in having no black on the clypeus, the yellow 

 mark on the vertex is absent, and the short black band on the 

 upper end of the first lateral suture of the synthorax is carried 

 vertically downwards to join the black line of the second suture, 

 forming with it a Y-shaped mark. 



Genus Burmagomphus Williamson. 



Genotype : B. vermiculatus Williamson ? nee Martin. 

 Species examined : B. pyramidalis, sp. nov. ; B. sivali- 

 kensis, sp. nov. ; B. sp. 



Rather small species (h.w. about 27 mm.). There is marked 

 parallelism between M 4 and Cu, in the fore- wing to beyond the 

 level of the nodus. As a rule only three rows of cells between 

 Cu 2 and hinder margin of hind-wing. (Individually there are 

 sometimes four rows). Generally only a single row of cells 

 between M, and 1\I,„ at distal end of pterostigma. 



The genus probably contains a considerable number of small 

 species which may shade off into Gomphus on the one hand and 

 into Platygomphus on the other. Ris has pointed out that the 

 parallelism of M t and Cu t is a feature common with Onychogomphus 

 the Javanese species, B. jacobseni Ris, has the apex of the ab- 

 domen very like that of a Platygomphus, though the venation is 

 definitely that of a Burmagomphus. 



The genotype is the species described by Williamson under 

 the name Burmagomphus vermiculatus (Martin). But his speci- 

 mens are very probably not conspecific with Martin's, in which 

 case the genotype would be without a name. 



B. pyramidalis, sp. nov., is closely allied to both Martin's and 

 Williamson's species ; as in the latter the oblique yellow stripe of 

 the dorsum of the synthorax is formed by the fusion of the 

 upper part of the dorsal, with the lower part of the juxta-humeral 

 stripe, a peculiarity which seems to occur in Martin's species as well. 

 The three may well be geographical races of a single species. On 

 the other hand the Bornean form which I described under the 

 name B. vermiculatus insularis must rank as a distinct species, it 

 appears to be related to a form from the Nilgiris represented in 

 the collection before me by a single very immature example. 



B. jacobseni Ris is quite distinct and as stated above in some 

 respects approaches Platygomphus. 



Lastly the new species, B. sivalikensis, from Dehra Dun is 

 again quite distinct, and possesses a complete juxta-humeral stripe. 

 In size and general proportions it resembles B. pyramidalis. 



