LUTE-HISTORY OF THE LxVDY-BIRD BEKTLE. 



479 



the larval or pupal stages. This phenomenon is not due to con- 

 finement or lack of foorl, as it takes place in the presence of 

 abundant food and in normal natural conditions. One can easily 

 watch what appears to be a form of Coccinellid degeneracy among 

 the larva? which frequent red and black currant bushes. I have 

 seen adult hipanctata and septemptmctata eat their own eggs in 

 boxes and on the trunks of trees, but never eat either larvae or 

 pupfe. This habit was also observed (6) in C. distincta in con- 

 finement. Palmer has, however, observed the imagines eating 

 larvje when there was no other food in the cages. On one 

 occasion I watched a hijninctata eat the abdomen of a small fly 

 caught in a spider's web. After a period of experimental star- 

 vation, two Adal'm hipimctata feasted upon a small weak relation. 

 Cannibalisni has also been noticed by Palmer, Johnson, Burgess, 

 etc. 



On hatching, the almost colourless larva? remain clinging to the 

 empty egg-shells, part (never all) of which forms their first meal. 

 They next fall upon one another. This same early habit of 

 cannibalism also occurs in Einlacliiia (27 ); the vegetarian Lady- 

 bird. If by chance an q^^^, lags behind the others, it never 

 hatches, for it is eaten by a larva which was born a little earlier. 

 The larvae remain on the egg-mass until fully coloured, and then 

 suddenly and simultaneously disperse. Approximately a quarter 

 of the larvpe are eaten before this dispersal. 



During the first instar there is great mortality, partly owing 

 to lack of sufliciently small aphis and pai'tly to cannibalism. The 

 young larva? do best in small boxes, as they seem incapable of 

 wandering far at this stage to find food, which, J. believe, is 

 chiefly obtained by good luck and persistent wandering. As 

 early as possible the larva? vv^ere isolated in separate boxes, but I 

 was inclined to think cannibalism in the early stages a physio- 

 logically highly suitable diet, and, with this in view, I fed the 

 hipunctata with small larvfe of other species. This practice was 

 not, however, always successful, as sometimes the eatei- became 

 the eaten. Partially successful attempts were made to feed the 

 larvae on cooked and raw hen's eg^. Coccinellid eggs were 

 always eaten when put in the boxes, but this is not a suitable 

 ])roceeding in the early instai-s, as an odd egg may escape and 

 iiatch. During 1918, 21 of '22 families wei'e reduced to one, as 

 the full significance of cannibalism was not then recognised. 



It is interesting to consider the intermittent competition 

 l)etween larvae of the same batch of eggs. It is not the larva 

 which first starts in the race which necessarily first reaches the 

 Avinning-posfc. It may be the first to reach the end of the first 

 instar, but during the short quiescent period which precedes 

 ecdysis, it becomes a defenceless prey to those who have lagged 

 behind. This special danger recurs at the end of each of the fonr 

 instar s. 



The method of attacking the pupa is always the same, the 

 la.rva beginning on the ventral side near the anal point of 

 attachment and eating its way upwaids and forwards, half 



