LYCiENINiE. 323 



at their food. The larvaB crawl slowly, and are somewhat slug- 

 like in their motions ; they appear to rely mostly for their pro- 

 tection on the similarity of their colours to their surroundings, 

 those of many species responding most remarkably to their environ- 

 ment. Of the larva, Chapman writes (in litt.) : " In the newly- 

 hatched larva, my material is inadequate to distinguish many decided 

 differences from those of Theclids and Chrysophanids ; Scudder's 

 table (see preceding volume, p. 317) is certainly quite erroneous. 

 The 7th and 8th abdominal segments are equally distinct in all the 

 groups from the 9th and 10th, which are more or less fused, at least, 

 difficult to distinguish (they tend to fuse in all- in the later stages). 

 I should certainly have said the head was proportionally larger (not 

 smaller) in Theclids if there was any difference. Two lenticles in the 

 region of iii are certainly very usual in Lycaenids, one in Theclids (but 

 Strymon pruni has two). Scudder's ' large, sunken, subcircular area,' 

 is the anal plate, and occurs very similarly in all groups on the 10th 

 abdominal segment, and has very much the same relations to the hairs 

 around in all. Concerning the adult larva, Scudder's table showing 

 the head to be smaller in Lycaenids, and prothoracic scutellum to be 

 obscured in them, seems, so far, to be confirmed. He notes, however, 

 that, in Theclids, the highest part of the larva is past the middle. 

 This is clearly erroneous, some of his own figures contradicting it. 

 It is somewhat true in some species, but the larva of Ruralis betulae, 

 with the short anterior, and long posterior, slope, is much to the 

 contrary. Theclid larvae retain, to the last skin, more or less definitely, 

 the two dorsal ridges or flanges. Lycaenids retain them, so that they 

 may be distinguished theoretically rather than practically, whilst in 

 Chrysophanids they are quite lost, or, at least, it takes some imagina- 

 tion to see them clearly. The Theclid larva is more or less angular, 

 as in the serrations of the dorsal ridges of Strymon pruni, depression of 

 scutellum in Thestor ballus and others, etc. Lycaenids are more or less 

 rounded, the only angularity being the occasional sharpness of the 

 lateral flange, whilst Chrysophanids tend to be very flattened (or to 

 look so) and slug-like. A feature that is extremely common in 

 Lycaenids, and does not, so far as I have observed, occur in the 

 other groups, is stellation or spiculation of the hair-bases (very 

 marked in Celastrina cur/iolus)." 



The connection of Lycaenid larvae with ants has already been 

 somewhat fully noticed in the preceding volume (pp. 30-37). This 

 connection was first noticed much more than a century ago by Esper, 

 and has now been observed in many Palaearctic and Nearctic, as well 

 as tropical, species. Many Lycaenid larvae, as has already been 

 noticed, have the power of exuding, from a special evaginable gland 

 on the dorsum of the 7th abdominal segment, at the solicitation of 

 the ants, a drop of fluid which the ants greedily drink. The associa- 

 tion of ants has been noticed with several species — Plebeius argus (aegon) 

 With Lasius niger (Aurivillius, Ent. Tids., v., pp. 190, 227), and with 

 L. alienus (Viehmeyer, in litt.), Plebeius argyrognomon with Formica 

 cinerea (Torka teste Viehmeyer, in litt., and Thomann, Beobach. Symb. 

 L. argus, etc., 1901, pp. 1-40). Agriades corydon with Formica niger, 

 and more rarely with F. rufa (Krodel, Allg. Zeits. fur Ent., ix., 

 pp. 103-4), also with F. flava (Ray ward, Entom., 1906, p. 197), 

 Agriades bellargus with Lasius niger and L. flavus (Rayward, Entom., 

 xxxix., pp. 219-220), Polyo mmatusicarus with L. flavus (R&yvt&vd, Entom., 



