BY R. J. TILLYARD. l95 



In carrying out this research, one of the principal difficulties 

 was found to be the extreme thinness of the wing-cases, through 

 which the slightest pressure served to disarrange the tracheae. 

 Many specimens were spoilt through this. The plan which I 

 finally adopted was to dissect ofi" the two wings of one side 

 together, under water, with a fairly large portion of the pleural 

 attachment; then, to separate the two wings by a sharp cut in 

 the plane of the wiyig-case, not transverse to it, so as to cut the 

 attachment in halves without exerting any pressure on the wing- 

 cases themselves; and, finally, to fioat each wing out separately 

 on to a slide with plenty of water, let down a cover-glass gently 

 on to it, and photograph it while still in the water. If desired, 

 the attachment can be cut completely away in order to obtain a 

 planer surface for focussing; but as this usually pulls the delicate 

 anal trachea away from its attachment to Cu, it should only be 

 done for enlarged studies of other parts of the tracheation, as 

 for instance, the studies of the bridge (Section B, above). 



Another great difficulty is the fact that some larvae have very 

 hairy wing-cases, while many have them very deeply pigmented. 

 The wing-cases of Argiolestes are very hairy and rather thick; 

 those of Isosticta nearly always jet black. Some ingenuity is 

 required in manipulating these insects. Remembering that the 

 dark colour of these larvae is mainly due to their being bottom- 

 dwellers, I placed some Argiolestes and Isosticta larvae in a jar 

 with a mass of water-weed alone, no sandy bottom being sup- 

 plied. After wandering round the glass bottom for some time, 

 most of the larvae took to the water-weed, on which they rested 

 clumsily with outspread legs (showing their inability to cling 

 closely in the way that habitual weed-dwellers do). One Argio- 

 lestes and several Isosticta larvae went through an ecdysis in this 

 position; and instead of rapidly darkening, as they usually do, 

 remained pale brownish for many days, showing even a tinge of 

 green also. This was what I expected. From these larvae, I 

 obtained very fair photographic results, from which the diagrams 

 (Plate xii., figs.7, 11) have been taken. I also obtained a photo- 

 graph of an Isosticta larval wing soon after ecdysis, and was 

 interested to notice that all the imaginal wing-venation showed 



