250 BRITISH LEPIDOPTERA. 



seems to be no difference in colour, marking, or form of wing. The 

 $ antenna is the longer, as about 17mm. to 14mm., and is also 

 thicker, as o'8mm to o^mm. The joints are about equal in number 

 (76), and the scaling is similar to that of other Eumorphids, viz.> 

 one row of very short and two of longer (one a little longer than 

 the other) scales to each segment. The first tibia is longer in the 

 c? , as 6'omra to about 5-5111111. ; the $ spur is about 3mm. in length, 

 with the comb extending for about two-thirds of this length, the spur 

 is proportionally very broad (i^mm.); the spur on the £ first tibia 

 is a little shorter (2 , 5mm.), and the comb extends along about half 

 its length. The fan which the $ carries on the 1st abdominal 

 segment (really the 2nd) arises just behind the spiracle, about the 

 middle of the margin of the dorsal plate, and there passes to it a 

 narrow process from the anterior margin of the ventral plate ; this 

 has, no doubt, something to do with the process of erecting and 

 expanding it. It is really a small (for the size of the moth) wisp of 

 hairs about 3'5mm. long. The following spiracles are not in the 

 dorsal plate, but really in the lateral membrane. The marginal 

 batons are obviously modified scales, being distinctly striated, though 

 having the fusiform pointed form they have in other species (Chapman). 



Gynandromorphism. — The following are the only gynandro- 

 morphs that we have been able to trace : — 



a. Imperfect gynandromorph. Wings on both sides equally long ; the 

 forewings 47mm. long, left 26mm., right 25mm. broad; a decided difference in 

 the hindwings. Markings and coloration of the wings somewhat different. Left 

 $ antenna not quite fully formed, right $ normal. Left frenulum perfectly ? , 

 the right one, half $ and half ? ; the little hook on the forewing for fastening the 

 bristle is wanting. Abdomen less pointed than in normal specimens ; segmentation 

 c? , being 7-ringed, the last two segments divided by a complete incision, and 

 coloured and marked as in the J 1 . On the left side a developed $ anal clasp is 

 present, but moved so out of place that it lies obliquelv over on the right ; the 

 corresponding clasp on the right side is rudimentary {cf. Spever, Stett. ent. 

 Ztg., 1868, p. 238). 



)3. Left c? , right ? . As regards sex a perfectly halved hermaphrodite. 

 Right forewing 48^111111. long, at the hind margin 26|mm. broad, left 46mm. 

 and 24mm. respectively. Coloration of the ? wings somewhat deeper. Right 

 antenna ? , le(t c? ; frenulum and form of the segments likewise. Of the 

 outer sexual organs only the large, somewhat projecting, lett anal clasp is 

 distinct ; on the ? side the clasp is wanting. Bred by Greutzenberg of Dantzig 

 (cf. Speyer, Stett. ent. Ztg., 1869, p. 237). 



y. A gynandromorphous example in Dr. Staudinger's collection [Staudinger, 

 in. litt. (Schultz, ///. Woch. f. E?it., ii., p. 393). 



Variation. — The species exhibits a certain amount of size 

 variation [British examples are often of good size, e.g., 4ms. ,4.1m. 

 at Hastings (Kent), 4f ins. at Chickerell (Richardson)], and sometimes 

 marked colour aberrations occur. Bartel notes (Pal. Gross-Schmett.^ 

 ii., p. 135) that examples from the neighbourhood of Constantinople 

 are distinguished by their large size and fineness, as also are such as have 

 fed chiefly on flowers of oleander; he says that Miitzell, of Berlin, bred 

 really wonderfully magnificent specimens by supplying the larvae with 

 this kind of food and raising the temperature as much as possible. In 

 rare aberrations the green of the upperside becomes yellowish-brown. 

 Unterberger, of Konigsberg, records (III. Zeits. fur Ent., iii., p. 232) 

 a colour aberration of a cf which had the beautiful olive-green of 

 normal examples replaced by olive-brown, becoming ochre-yellow 

 on the underside of the hindwings. The specimens from the Hawaiian 

 Islands are stated by Butler (E.M.M., xiv., p. 47) to differ in no 



