SOUTH AFRICAN CATERriLLARS 193 



groups of such as appear to be alike and put those of 

 each group into a separate breeding cage. These breeding 

 cages receive successive numbers, say 1, 2, 3, 4, etc. For 

 each breeding cage a separate file is kept into which are 

 put all observations relating to this breeding cage. Of 

 the caterpillar a detailed description is made as to colour, 

 pattern, and at least one specimen is preserved for study 

 of structural characters; another one is blown for a 

 cabinet specimen. The cage is looked over from day to 

 day, and a daily record kept. At every inspection a care- 

 ful search is made amongst the frass and excreta for cast 

 skins, and, when found, these cast skins and headcases 



are preserved and labelled B.C. No — , with date, 



Wlien there are several caterpillars in one cage we find 

 that they do not all moult simultaneously as a rule, and at 

 least one which is on the point of moulting, as also a 

 newly moulted larva should be preserved, so as to get 

 representatives of the beginning and of the end of an 

 instar. A couple of days after the moult another larva, 

 which has then obtained the full colour pattern and 

 armature characteristic of the instar, must be preserved. 

 Again, before preserving make a detailed description of 

 the larva. 



If this plan be carefully carried out throughout the 

 various larval instars and the pupa stage (specimens of 

 newly-formed and mature pupa having been preserved), 

 we ultimately get a complete record of the immature 

 stages of the insect, showing: 



(a) dates of moulting, and thus number and duration 



of instars, 

 (&) colour pattern of the larva in its various instars, 

 and thus the changes occuring with each succes- 

 sive moult, 



(c) notes on the habit, food-plant, etc., of the cater- 

 pillar, 



(d) specimens of the larva in its various instars for 

 study of the relative position of setae, armature, 

 etc., 



