﻿28o LÉPIDOPTÉROLOGIE COMPARÉE 



Icave the Gentian, and the segments are markedly bulging, i. e. 

 the incisions are marked, and also each segment has on each 

 side two definite fulnesses or prommences about spiracular 

 level, a larger one above, towards anterior margin of segment, 

 and across the dorsum is another stiU larger along the posterior 

 half of each segment. 



Sept. i8th. Condition much as yesterday. 



Sept. içth. Much the same, larvae measure fully 5 mm. if 

 not perhaps 5.5. It is observed that when the disturbance of 

 letting m the hght to see what is going on leads the ants to 

 move the brood to another place, tlie larvae of alcon are usually 

 the hrst to be seized and carried off, before the ants own brood 

 is touched, the movement is very simultaneous and several ants 

 begin it, so that some ant brood goes off pretty much with the 

 (lie on. When the thing is donc half-heartedly and only some 

 brood is removed, the alcon are nearly always taken first, and 

 hardly ever left behind. 



Sept. 2oth. Much the same, the alcon are usually at the margin 

 of the heap of brood, not in or under it as arion usually was. 

 Hâve not been able to see alcon eat any ant larva but to-day 

 saw an ant mouth to mouth with one, just as they may be seen 

 with their own larvae, no doubt when feeding them. Twice 

 to-day noticed how an alcon larva was carried off on distur- 

 bance, before any ant larva was touched. 



Sept. 22nd. Seen to-day in one nest in middle of brood. 

 The segments are swoUen so as to make the latéral outline a 

 séries of curves with the incisions well marked. 



Sept. ^oth. The larvae of alcon remain much the same, over 

 5 mm. long with swollen segments. The larvae of the ant 

 diminish in numbers and amongst the contents of " midden 

 are remains of ant larvae consisting of the collapsed skin and 

 dark intestine with contents, as though ail their fluids had been 



