﻿LÉPIDOPTÉROLOGIE COMPARÉE 283 



probably accounts lur ant larvae being found in the " midden " 

 dead but only partially sucked out. 



Oct. iSth. Receivcd larvae ol U . unibratiis last evcning irom 

 Mr. Donisthorpc, placcd two of tlicm m nests 4 and 5 and larva 

 of alcon from 4 amongst the mass ol iiiiibralus larvae in a tuVje. 

 The M. scabrinodis approachcd the unibratus larvae in their 

 nests, looked askance at them and shortly not a few approached 

 them and seemed to seize them or to be about to seize them, and 

 instantly drevv back sharply, ahnost as if stung, or animated by 

 extrême disgust, probably not the former as they nnmcdiately 

 resumed an ordinary demeanour. 



Oct. içth. The L. alcon restored to nest 4, and was at once 

 accepted, its 24 hours sojourn with F. iniibraius larvae did not 

 seem to hâve given it any of the attributes of those larvae in the 

 view of its hosts. In the uinbraius tube it had donc nothing but 

 was possibly less stout. 



The unibratus larvae in the scabrinodis nest are uninjured, 

 but cast out as midden matenal. 



Oct. 21 st. The L. alcon seem much the same, quite as stout, 

 5-6 mm. long, but do not grovv further, collapsed larvae of sca- 

 brinodis are regularly found m nests (5 yesterday), but never 

 m nest 5, that has no alcon guest. 



In No. 4 nest one larvae of lanbratus is seen bitten in several 

 places, the other and those in 5 are not seen, whether demolished 

 or somewhere hidden is not évident. 



Put two more unibratus larvae in nest 4. 



Oct. 2yd. Fmd j larvae of unibratus in nest 4 and the same 

 in nest 5, discarded from nest and lymg about, each separate. 

 Ihere is no definite midden in either of thèse nests and possibly 

 thèse larvae are toc bulky to be taken to one, which might hâve 

 been instituted by using thèse larvae so. A definite midden is 

 interfered with by my process of raiding it for damaged ant 

 larvae. 



