458 SUMMARY OF CURRENT RESEARCHES RELATING TO 



tion are against it. The study of halved gynandromorphs and inter- 

 sexes is against the idea that the secondary sex characters are in any 

 way dependent on the gonads or on any portion of the internal sexual 

 apparatus. There seems little likelihood of there being some gland of 

 internal secretion independent of the sexual organs. 



It may be that the tissues are sexually differentiated from the first in 

 insects and require no secretory stimulus, but stylopization and tempera- 

 ture experiments show that the secondary sex characters can be changed 

 at a late stage. It may be that there is in all animals a sexual formative 

 substance which in some cases requires the co-operation of internal 

 secretions if the sex characters are to be perfected, and in some cases does 

 not. But on this view, unless the inferences from gynandromorphism 

 are incorrect, the sexual formative substance must be produced in various 

 parts of the body, if not by all the tissues. 



Digestion in Insects.* — Eldon W. Sanford confirms the observation 

 of Petrunkevitch that fat is digested and absorbed in the crop of the 

 cockroach. A. Krausef finds that in some carnivorous insects some 

 digestion takes place in the crop and gizzard. 



Habits of Larva of Lycsena arion.J — T. A. Chapman discusses the 

 inter-relations between ants and the larvse of LycEenines. The initial 

 point would probably be found in the case of those " blues " whose 

 larvte are collected by the ants and placed on food-plants actually on the 

 ants' nest, as occurs with argyrognomon, roridon, and hellargus, and 

 probably other species. Perhaps the nest is extended towards a suitable 

 plant. 



" It might ])e said that the ants do take these larvge to their nests, 

 as they leave them, when at rest, on the rootstocks of their food-plants, 

 and cover them with loose material ; and I have seen ants remove a 

 small larva that they appeared to think was in danger when I disturbed 

 them. A slightly deeper enclosure in the nest, and a cannibal proclivity 

 on the part of the larva, would initiate the arioii habit." 



The normal number of moults in Lyc^nines is four, and the normal 

 instar for hibernation (of those that hibernate as larvas) is the third. It 

 is probable that avion was at first carried into the nest in the third 

 instar, and there hibernated harmlessly. 



Some carried in might take another moult before hiberuating and 

 begin eating the larvse of the ant. A tendency to vary towards 

 diminished size would also be involved in the causes leading to hiberna- 

 tion in the fourth instar. Those that survived the animal diet and the 

 dangers from ants would benefit by the protection during the winter 

 and the assured dietary. But the most remarkable fact as regards the 

 hibernation of L. arion is that it does so as a half-grown larva in the 

 last instar. No other Lycajnid is known to do so. This change of 



* Proc. Soc. Exper. Biol. Med., xiii. (191G) p. 193. See Physiol. Abstracts, i. 

 (1916) No. B, p. 360. 



t Zeitschr. Allg. Physiol., xvii. (1916) pp. 164-7. See Physiol. Abstracts,!. 

 (1916) No. 8, p. 360. 



t Trans. Entomol. Soc. London, 1916, parts iii.-iv. (publ. 1917) pp. 815-21. 



