BY R. J. TILLYARD. 105 



The cocoons were opened with a dissecting needle, by scratching 

 the sand away along a zone corresponding with that along which 

 a boiled egg is usually opened at breakfast. The appearance of 

 a shiny black inner envelope was a sign that the larva had been 

 parasitised, this being the cocoon of the Chalcid. The Eriocrania 

 pupa? were free inside the cocoon, and could easily be lifted out 

 without damage, on the point of the needle. They were all very 

 inert, and made no attempt to move. I therefore killed and 

 dissected them straight away, and was pleased to find that the 

 removal of their wings was a very simple operation. The wings 

 are even less glued together than is the case with the Mecoptera 

 and Planipennia, being only slightly joined at their apices. 

 These were at once separable with a needle, while the bases were 

 easily cut away from the thorax with a sharp dissecting-knife. 

 The separate wings were then floated out on to a slide, and ex- 

 amined first of all without a cover-slip. At this stage, drawings 

 were made with the camera-lucida. 



The tracheation being so fine, a high power was required for 

 the study of the basal specialisations. For this, it was necessary 

 to let down a cover-slip upon the wings. This had to be done 

 very carefully, for fear of disarranging the delicate tracheation; 

 but in each case the operation proved successful; so that further 

 drawings under a higher power, together with photomicro- 

 graphs, could be taken. 



The wings being different in certain important particulars, in 

 spite of their homoneurous appearance in the imago, it is advis- 

 able to deal with each wing separately, and then to make a com- 

 parison between them. 



T h e F o r e w i n g. (Plate iii., figs.1-3; Text-figs. 1-2). 

 The tracheation of the forewing is of a very generalised type, 

 as may be seen from Text-fig. 1. All the main tracheae are quite 

 distinct and separate from their bases outwards. The points to 

 be noted are: — 



(1) K, gives off' a pterostigmatic veinlet R llt , homologous with 

 one of the corresponding veinlets in many Mecoptera, Megalo- 

 ptera, Planipennia, and in a few Trichoptera. 



