360 AXTS. 



that the ant is soon in a bad way and has quite enough to do in attempt- 

 ing to free himself of his encumbrances without taking any further 

 interest in the butterfly, from which he retreats as well as possible. It 

 is exceedingly ludicrous to observe the ants endeavoring to free them- 

 selves, their legs move awkwardly and their mandibles are opened and 

 closed in evident annoyance and perplexity, and they are also much 

 concerned about the state of their antennae for the obnoxious scales will 

 not be shaken off, and they seem to become very low-spirited." 



Another Indian Lycsenid (Allotinus horsficldi) attends aphids in the 

 butterfly stage, stroking them with its long fore-legs and sucking up 

 the excreted honey-dew with its proboscis (Bingham, 1907)! These 

 observations suggest the existence of complicated " three-cornered " 

 relations between ants, Lycsenids and aphids. It would be interesting 

 to know whether the Allotinus larva is attended by ants in the same 

 manner as the aphid is attended by the butterfly. In the case of 

 Fcniseca tarquiniiis I find a triple relationship of a different kind, for 

 I have observed the alder blight, on which the larvae of this Lycaenid 

 feeds, being attended by at least six different species of ants of the 

 genera Prenolepis, Camponotus, Formica, Cremastogaster and Dolicho- 

 dcnts, so that here we have a condition very similar to that described 

 by Silvestri (ride supra, p. 357) for Tapinoma, Tettigometra and 

 Hyperaspis. 



It will be seen that the intimate relations described in this chapter 

 as existing between ants on the one hand and the various Homoptera 

 and Lyc?enid caterpillars on the other, excepting, of course, such forms 

 as Spalyis and Fcniseca, have a common peculiarity. In all of these 

 cases the ants are supplied with food in the form of an excretion or 

 secretion elaborated from the juices of plants. Wasmann has therefore 

 designated these relationships as trophobiosis to distinguish them from 

 the cases of myrmecophily proper, which will be the subject of Chapters 

 XXI and XXII. 



