120 POLYOMMATUS (LYC^NA) AGESTIS. 



merit, on a bit of linen, the second followed in the 

 same way on the 16th, and on the 21st and 24th they 

 changed to pupse. The two remaining larvse soon after 

 fixed themselves, but died unchanged, probably the 

 effect of insufficient food. 



The newly-hatched larva is very minute, with a glis- 

 tening blackish head, stoutish body, of a light drab- 

 green colour, velvety and hairy ; its size is doubled in 

 eight days, and when a month old it is of the usual 

 Lyccena-shsupe, one line in length, thick in proportion, 

 with small retractile head, the body of a dull pinkish- 

 brown colour, with darker dorsal stripe, and rather 

 hairy. 



On waking up in spring it is of a dingy slaty-green 

 colour, and early in March it moults, when the old 

 skin is left attached to the plant like an empty shell, 

 not in the least shrivelled, but split open laterally along 

 the ridge above the legs. The larva now becomes quite 

 pale green on the back, broadly pinkish along the lateral 

 ridge, and still hairy. Early in April it is nearly an 

 eighth of an inch long, of greenish flesh colour, palest 

 on the second segment and dorsal eminences, pinkish 

 in the dorsal hollow, and also beneath the spiracular 

 region, the long whitish hairs closely resembling those 

 of the food-plant. 



The last moult occurs about the 21st of April, when 

 it is three- sixteenths of an inch long, and attains its 

 full growth of barely half an inch early in May ; during 

 this interval of course the larva shows all its charac- 

 teristic details, which are just like those of the local 

 Northern variety (Artaxerwes) (as described 6 Entom. 

 Monthly Mag., ? vol. v, p. 176), and all I can say of the 

 type form is that the green colour is more lively and 

 full, and the pink along the lateral region is darker, 

 inclining to purplish. 



It only remains for me to state that my experiments 

 have proved to me the truth of what Zeller kmg ago 

 suspected, and siuce then Newman and others have be- 

 lieved that Artaxerxes, Salmacis, and Agestis (Medon) 



