THE FRIEXnS AND ASSOCIATES OF CATERPILLARS. 963 



blue butterflies, tlie gland wliicli secretes the sugary fluid is not confined 

 to the caterpillars of these butterflies, but is also found in many of their 

 immediate allies, namely, in most of the hair-streaks or Theclidi, and in 

 one at least of the coppers, viz., Tomares ballus of Europe. Two ex- 

 planations readily ofter themselves : one, that in caterpillars so little 

 known as are these, it may v»-ell be true that the association of the ants 

 with the caterpillars has escaped notice ; or, on the other hand, that in 

 these instances the srlands secrete a fluid which has no saccharine ingre- 

 dients. Tiie advantage that it may be to the caterpillar to secrete a sweet 

 fluid attractive to ants is obvious, since the ants imdoubtedly keep oft' 

 many iciineumon flies and other enemies of the caterpillar, and the mutual 

 benefit conferred by ant and caterpillar is unquestionable. It is doubtful 

 if in the other cases the gland ever secretes a fluid having an offensive 

 quality which might equally serve as a protection against intruders, since 

 this means of defence is probablj' found, in caterpillars of this group, in 

 organs of a very different character upon the succeeding abdominal segment, 

 and it is hardly to be presumed that two organs, distinct in their position and 

 structure, should arise in one and the same animal for precisely the same 

 object. The use, therefore, of the median gland presumably not possess- 

 ing a saccharine character is very problematical. 



It should not be overlooked in connection with this subject that these 

 caterpillars are themselves fond of saccharine matters. They, and only 

 they, of all our butterfly caterpillars, attack flowers where honey is 

 secreted, and there is at least one form described in this work which has 

 come to have an entirely carnivorous diet, feeding upon plant lice bodily 

 for their juices, very likely for the same reason that the ants seek their 

 secreted fluids in detail. So, too, and possibly for the same reason, these 

 creatures not unfrequently show a cannibalistic tendency, feeding upon the 

 bodies of their own brethren when they are in a helpless condition, as pre- 

 vious to pupation. All these subjects are closely related to one another, 

 and need to be investigated more carefully in order to a complete solution 

 of their meaning. 



It is a curious thing that among the Lycaenidi these glands are found 

 on some species while not found upon others closely allied ; their occur- 

 rence in many members of the other two tribes of the Lycaeninae, 

 together with the impossibility of their independent origination in diflferent 

 genera render it probable that these glands first arose as long ago as 

 before the differentiation of the three tiubes of Lycaeninae. The brother- 

 hood of the ants and the caterpillars may therefore be of great antiquity. 



•»* See especially a paper by Mr. W. II. Edwards, ou the larvae of Lye. pseudargiolus and 

 attendant ants (Cau. ent., x: 131-136) and his account of the same insect in his Butterflies of 

 North America, "Vol. ii. 



