igi7.] F. F. IvAiDiyAw : Indian Dragonflies. 35 



The female differs from the male in colour, mainly in that 

 the markings on the head and dorsum of the thorax are of a 

 yellowish green colour instead of being bright blue. The large 

 mesothoracic triangle, so conspicuous in the male, is outlined in 

 colour but its central part is black. The abdomen has an antero- 

 lateral stripe on segments 2, 3, as well as a posterolateral spot of 

 yellow colour on the same segments. The latter mark only is 

 present in the male, at least when adult. The rest of the abdomen 

 in both sexes is black. 



The anterior surface of the two posterior pairs of tibias of the 

 male is white and the curious exudation, referred to by de Selys 

 {loc. cit.), is well seen in the spirit specimens. 



Group Unimacui,ata. 



24. Rhinocypha unimaculata, Selys. 



(PI. IT, fig. 3). 



R. unimaculata, Kirby, Cat Odonata, p. 113 (1890). 



This is the largest of all Indian species of the genus. It is 

 abundant from Darjiling to Assam, but the limits of its range are 

 not known. 



The dimensions of a male specimen are : — 



Length of hind-wing . . 31 mm. 

 ,, ,, abdomen . . 24 ,, 



25. Rhinocypha trimaculata, Selys. 



R. triinaculata, Kirby, Cat. Odonata, p. 113. 



The locality ' Thibet ' given for this species by Selys should 

 almost certainly be rather Assam. I have not seen an example of 

 this, one of the smallest and probably one of the rarest of the 

 Rhinocyphas ; it is closely allied to R. ignipennis. 



26. Rhinocypha ignipennis, Selys. 



R, ignipennis, Kirby, Cat. Odonata, p. 113. 



Williamson, Pi'oc. U.S. Nat. Mus., XXVIII, pp. 179-181. 



2 & cf {^%^-], Cherrapunji, Assam, i a* , r 2 (-|§-), Shillong, 

 Assam, 4,900 ft. 



The female specimen has the anterior surface of the femurs 

 marked with white. 



Group Fenestrella. 



This group is restricted entirely to the mainland of Asia 

 ranging from the Himalayas to the extremity of the Malay Penin- 

 sula. The species or races present marked individual variation, 

 and in view of this I think it well to treat some of the species as 

 local races, a course already suggested by de Selys in his paper on 



