35 s ANTS. 



median gland a glistening droplet of a colorless and presumably sac- 

 charine liquid which is eagerly imbibed by the attendants. The 

 function of the tentacles is unknown. De Niceville believes that they 

 may represent vestiges of a pair of organs which are very long and 

 tufted in an Indian Lycsenid (Curetis thctys) and are very rapidly 

 whirled around when the caterpillar is touched, as if for the pur- 

 pose of frightening away its enemies. From the fact, however, 

 that the tentacles of the European and North American species are 

 most frequently protruded when the caterpillars are being caressed 

 by the ants, Thomann concludes that these peculiar organs act as signals 

 and diffuse some odor which serves to attract and fascinate the ants. 

 Xot only are many of the Lycsenid larvae attended while feeding on 

 the leaves or flowers of plants, but they are often found in the ants' 

 nests. Some of the species pupate and even hatch as butterflies in the 

 galleries of the nests. Only in a few instances have the names of the 

 attendant ants of our northern Lycaenids been recorded. As would be 

 expected, these are members of the genera Lasius, Formica, Campo- 

 notits, Prcnolcpis and Mynnica. In India, according to Dodd (19020 ), 

 Thwaites (1881) and de Niceville (1900), species of Phcidolc and 

 Cremastogaster, and especially (Ecophylla smaragdina, are the principal 

 attendants, ffi. smaragdina, in that country and in Australia, is, in 

 fact, constantly found with many species of the caterpillars and often 

 keeps them in the silken nests and " cow-sheds " described in a previous 

 chapter (pp. 223,224). De Niceville says that the butterflies will often 

 oviposit only on trees inhabited by ants: "If the right plant has no 

 ants, or the ants on that plant are not the right species, the butterfly 

 will lay no eggs on that plant. Some larvae will certainly not live 

 without the ants, and many larva? are extremely uncomfortable when 

 brought up away from their hosts or masters. In many cases it is just 

 as important for breeding purposes to know the right species of ants 

 as to know the right food-plant. In Kanara this is particularly notice- 

 able in the cases of Castalhts ananda, de Niceville, Zcsiiis chrysoiiwllus, 

 Hubner, Aphnccits lohita, Horsfield, and Catapoccilma clcgans, Druce. 

 On one occasion Mr. Bell was collecting larvae at Katgal, and the ants 

 were principally on Zizyplnts rugosa, Lamk. (Natural Order Rham- 

 neae), but were also swarming all over six or seven species of different 

 trees all round, and on all of these trees there were larvae of C. ananda 

 covered with ants and eating the leaves of the trees in all cases. Since 

 then Mr. Bell has noticed the larva of C. ananda eating the leaves of 

 many different plants and always in company with the same species 

 of ants. With regard to the Zczius, Aphnccns and Catapoccilma men- 

 tioned above, the female butterflies first look for the right species of 



