122 KRNEST HILL AND L. G. HAYDON. 



we have an imagO; herein referred to as Pyretophorus 

 cinereuSj Avhich is much larg^er than the largest funesta, 

 and by its naked-eye appearance is obviously different, much 

 more than can be shown by any detailed description (PI. XXV, 

 fig. c). The larva of this comes near in resemblance to 

 thiit of funesta, but the position of palmate hairs and the 

 shape at termination are different (Pis. XY, XVII, fig. b). 



Variations in size are to some extent due to external cir- 

 cumstances. Nearly all our specimens were obtained from 

 larvee kept in basins until the insects emerged from pupal 

 stage. Those which were young and small when captured, 

 more particularly in cool weather, when larval existence was 

 prolonged to seven weeks, developed generally into smaller 

 insects. Their growth was stunted, and they suffered from 

 debility, as a result of unnatural conditions. It was observed 

 that after some weeks' captivity the metamorphosis into pupa 

 took place, when the larva appeared still some way short of 

 full grown ; the pupa was smaller, the imago also, which 

 experienced much difficulty in getting free of pupa skin, 

 often losing hind legs, and not seldom life, in the process. 

 This did not happen to larvaa which were full-grown at time 

 of capture, unless shaken up in travelling. 



We have identified the larva of nine of the fifteen species 

 enumerated on page 112. The inost certain method of 

 identifying larv^ of a species is to breed from the imago and 

 watch the larva from egg* to final metamorphosis. This 

 proved impracticable, because conjugation did not take place 

 in captivity, and impregnated females were not captured when 

 required. Larvas were collected, examined alive, sorted into 

 groups, watched to metamorphosis, and specimens preserved 

 from time to time, and last larval moults mounted. With a 

 little practice familiarity is gained and distinction between 

 different species readily appreciated. The characters of each 

 species are sufficiently defined to leave little room for error if 

 sufficient care is taken. 



The points of importance for differentiation of larva? have 



