LARVJE OF POLYOMMATUS ICARUS AND THEIR CONNECTION WITH ANTS. 109 



secretes a fluid very attractive to ants ; in the hope, therefore, that 

 they may be of interest, I give some particulars of an experiment con- 

 ducted by me a few days ago with two larvae of icarus and a worker 

 ant of the species Formica flava, which not only proved the existence 

 in these larva? of a highly-developed secretory gland, but also showed 

 a high order of instinct, or intelligence, in the ant. 



The larvae experimented with were reared from ova deposited, in 

 confinement, on Lotus corniculatus, by a $ taken by me at Folkestone 

 in August, 1906 ; the ova hatched on the 26th and succeeding days of 

 the same month, and the larva?, after growing satisfactorily for three 

 or four weeks, ceased feeding, and laid up for hybernation about 

 September 2] st. 



They were wintered on a growing plant of Lotus corniculatus in the 

 open, and, during the second week of March of this year, some of the 

 survivors were removed to a warm room and placed upon cut sprigs of 

 Tri folium repens ; this food they commenced to eat during the follow- 

 ing week, and one of them moulted on March 26th. An examination 

 of this freshly moulted larva revealed the fact that it possessed a long, 

 narrow, transverse gland on the dorsum of the 7th abdominal segment, 

 similar to that possessed by the larva? of our British Lyca?nids — 

 arion, corydon, and bellargus — and as, in addition, the two conspicuous 

 whitish evaginable tubes on the next following — i.e., the 8th — segment, 

 carried by both corydon and bellargus, were also present, it appeared 

 probable that icarus might, like its allies, have some connection with 

 ants. As soon as opportunity offered, therefore, I put the matter to 

 the test, and was soon satisfied that icarus is endowed with a gland 

 which yields a fluid probably similar in character to, and equally 

 attractive to Formica flava as, that produced by arion, corydon, and 

 bellargus. 



The demonstration was obtained in the following manner : — The 

 larger and more vigorous-looking of the two larva? employed, was first 

 placed upon the stage of a small dissecting microscope carrying a lens 

 giving a magnification of 16, and an ant was then persuaded to walk 

 upon the stage from a pill-box by means of a camel-hair brush. The 

 ant appeared somewhat sluggish, and at first evinced no interest in the 

 larva, but, after a few moments, walked slowly over its back, and pro- 

 ceeded to clean itself, and particularly its antenna?, very thoroughly. 

 This operation lasted some four or five minutes, and, during its con- 

 tinuance, the larva remained perfectly quiescent, showing no sign 

 whatever of irritation or resentment at the presence of the ant, and 

 careful examination of the gland and evaginable tubes disclosed no 

 sign of movement in these organs, so that I began to fear that the 

 gland might prove, after all, to be a " blind " one, or at least im- 

 perfectly developed. Suddenly, however, after having completed a 

 most elaborate " toilet," including the careful cleaning of its antenna? 

 by passing them many times between the prothoracic legs, the ant ran 

 straight to the gland, and, in the most methodical and business-like 

 manner, began to caress it with the antenna?, evidently well aware of 

 the delectable drink waiting to be yielded for the asking, yet the result 

 for several minutes longer continued disappointing, for, although the 

 gland showed some slight sign of contraction, followed by distention, 

 there was no indication of the protrusion of the inner membrane, 

 which always precedes the flow of fluid in corydon and bellargus. 



