RHAGADES GLOBULARIJE. 409 



species, the antennal shaft light greenish-blue, with the pectinations 

 very dark brown, and much less serrated than in A. statices. From 

 the neighbourhood of Constantine, in May, etc. Staudinger queries 

 (Cat., p. 44) this as being identical with soror, Eamb., Cat. Lep. 

 And., p. 187 (1866), which came from Granada, and is diagnosed as 

 " viridi-aenea vel cuprea ; antennis gracilibus, dentibus brevioribus," 

 and is said to differ from R. globulariae in the more slender antennae, of 

 which the more pointed pectinations are only half as long. Staudinger 

 is inclined to consider it as identical with var. notata, Zeller. 



y. ? var. incerta, Staud., " Stett.Ent. Zeit.," xlviii., p. 72 (1887).— Size 20-25 mm. 

 The pectinations lie close together, so that they appear very thin. The antennae 

 of the J are scarcely serrated (at least less so than in E. globulariae). The fore- 

 wings are blunter and darker green than the very glossy tint of R. globulariae, and 

 similar to that of R. solana. A single female (from Margelan) is almost blue-green, 

 with more gloss than usual. The black-green fringes contrast moderately with the 

 green wings. The hind-wings are transparent-black, as in R. globulariae, and on the 

 underside they are sometimes (especially in the Margelan specimens) rather strongly 

 tinged with green. The fore-wings beneath are without any green-blue tinge. The 

 head, thorax, and ventral side of abdomen are of a rather glossy green tint, the 

 latter, above black, tinged with green. 



Staudinger notes this insect in " some numbers from Margelan and 

 Namangan, also a single individual from the Alexander mountains, 

 and thinks it may be, perhaps, only a very small form of R, globulariae, 

 as it has almost the same form of antennae." But he further states 

 that " this much smaller incerta, in spite of its almost identical 

 antennal form, looks so different from R. globulariae (more especially 

 from its much blunter fore- wings), that I cannot consider it a local 

 form thereof, although that is possible. On the other hand, the insect 

 appears so similar to large R. solana, that I can only distinguish them 

 by the antenna?, which appear much thinner in incerta, and probably 

 are so ; the antenna? of J solana are also more serrate than those of 

 incerta. As I have received both insects from Margelan and Namangan 

 in some numbers, and can pretty readily separate them by the antenna?, 

 1 am inclined to think they are distinct species." 



Egg. — The egg of R. globulariae is, according to Buckler, of a long- 

 oval shape, about -^ " long, and yL " wide, having at first a depression 

 on some part of the surface, and adhering lengthways to the substance 

 whereon laid, singly, or sometimes two or three together. The shell 

 is very finely ribbed, and of a deep yellow colour, which changes a 

 few hours before hatching to a dull pinkish, or to a light brownish 

 tint, showing a deeper brown spot at one end, and by that time the 

 egg has become very plump. 



Habits of larva. — The young larva leaves the egg about the 

 middle of July. It is a small yellow mite, and is very active until 

 it meets a suitable spot on a leaf, when it at once begins to mine, 

 boring into a leaf of Centaurea nigra, embedding itself in the substance, 

 and forming a semi-transparent spot, which is afterwards enlarged 

 into a small blotch between the upper and under epidermis of the leaf, 

 excavating the leaf almost to the tip. When necessary the larva leaves 

 one leaf and enters another, the vacated blotches, where the parenchyma 

 has been eaten out, becoming somewhat conspicuous. The first moult 

 occupies nine days, according to Buckler, and the three succeeding 

 moults follow after about four clays feeding and nine days of rest on each 

 occasion, the larva leaving its mine and spinning a silken web, on 

 which it rests whilst the operation is performed. When mining, the 



