UY R. j. TlLLYARD. 447 



c. Outstandi>iy fentures of colour-development. — -I doubt very 

 much if these caii ever be used with advantage for generic pur- 

 poses. The brilliant postocular spots that are develof)ed in many 

 of these species are a case in point. Not only in dried specimens, 

 but also in very immature or very old siDeciraens freshly cauglit 

 (especially in tho.se specimens v^^here pruinescence supervenes with 

 age and obliterates the colour-pattern), this particular character 

 is exceedingly unreliable. Again, genera based purely un 

 differences of colour ai-e strongly to be objected to; for instance, 

 Pyrrhosoma and Erythromma, and even Nehalleviia and Ayriou 

 (in which similarity or difference in the colouration of the sexes 

 is the distinguishing character) seem to me to l)e purely artificial 

 genera, and ought not to be admitted. One might as well elevate 

 Ischiiura aurora Brauer, into a separate genus on account of its 

 red colouration. 



iii. CotLs taut iinisexual. characters. — If these must be used at 

 all, one would prefer some male character, since, in all collec- 

 tions, the males are so much more numerous than females, and 

 many rare species are represented only by males. One of the 

 chief characters used by de Selys, is the presence or absence of 

 the ventral apical spine on segment 8 of the female, 1 doubt if 

 this is of as great importance phylogenetically as de 8elys 

 imagined; and as it does not appear to be correlated with any 

 male character, the objection to its use is obvious. I should 

 prefer to use it only when all other tests have been exhausted. 



I believe that, on the above lines, a good working classification 

 of this legion could be obtained. But I do not propose to attempt 

 it here, as it would be iifcessaiy to study every known species 

 from an asthetioyenetic vieiv-poiiU, in order to succeed. However, 

 I ofier a classification of the Australian species of the legion, 

 based on these lines, as follows : — 



'Inferior sector of triangle reaching back l)asally 



to beyond tlie basal pustiKjdal nervule 1. 



Inferior sector of triangle reaching back basaliy 

 onli/ at the moft as far as tiie basal poslnodal 

 nervule* '2. 



'Except in the furewings of Xanthocneniis, n.g., in which it reaches 

 just beyond. Thi^s genus is really the link between 1 and 2, but is most 

 conveniently placed in 2. 



