296 BRITISH LEPIDOPTERA. 



and Echidnae as stirpes {antea, pp. 266 — 267). Hiibner's action, there- 

 fore, must be considered as altogether ultra vires, as affecting Saturnia, 

 although his restriction of Heraea to carpini as type, in 1806, may 

 be held to have settled this to the exclusion of pyri and spini, tau 

 already being eliminated as the type of Aglia. 



Packard considers (Proc. Amer. Acad. Arts Sci., 1893, p. 57) that," in 

 the European Saturnia carpini and its allies, and the Pacific coast 

 species, Saturnia mendocino* and S. galbina, we have perhaps the 

 most generalised and primitive members of the family. In the larva 

 of S. carpini the setiferous tubercles are of the same size and shape 

 on the abdominal as on the thoracic segments, there being no 

 differentiation in shape, and size, or colour f, such as occurs in all the 

 other genera, except that the second and third thoracic dorsal 

 tubercles bear one or two bristles much longer than those on 

 abdominal segments 1-7, and about as long as those on the 8th 

 abdominal segment. There are six tubercles on the 8th segment, 

 being the same number as on the seven segments in front ; on 

 abdominal segment 9 there are four tubercles, and two on the 10th, 

 i.e., the suranal plate. The same number of tubercles on the 

 8th abdominal segment also occurs in Saturnia mendocino of 

 California." Packard further considers that the generalised form 

 of Saturnia larva is clearly shown by the fact that, in Platysamia 

 cecropia and all the other more specialised and later genera, there 

 are only 5 tubercles on the 8th abdominal segment, those 

 corresponding to the middle ones of Saturnia having, probably 



during embryonic growth, coalesced Meanwhile it 



is not unreasonable to suppose that all the more specialised genera 

 must have been derived from a Satumia-Yike ancestral form, i.e., 

 a larva of cylindrical shape, with all the tubercles, whether thoracic 

 or abdominal, of the same size, shape, and colour on all the 

 segments ; those on the 8th abdominal segment being of the same 

 number (six) as on the segments in front." 



Standfuss has carried out elaborate crossings of the species of 

 Saturnia, his earliest results being detailed (Handbuch der palaeark. 

 Gross-Schmett., pp. 66 et seq.) in 1896. The experiments on 

 which these results were based were carried over some ten years, 

 the parent species dealt with being Saturnia pavonia, S. spini and 

 S. pyri. The various primary hybrid forms are described in detail 

 in this work as follows : 



I. S. hybr. bornemanni, Stdfss. (pavonia S X spini ? ). — The eggs first 

 laid were regularly deposited, and were fertile to the extent of from 60 to 85 per 

 cent. The larvae in all five stages bore a much closer resemblance to those of 

 S. spini than to those of S. pavonia, though this became less pronounced after the 

 second stage was passed. The cocoon and pupa were both intermediate in structure, 

 but the perfect insect was nearer to S. spini than to S. pavonia. This applies to 

 both sexes, but is more easily seen in the male, the males of the two parent species 

 differing (as is usual) more than the females {loc. ctt., pp. 66 — 74, tigs. 1—3, pi. ii., 

 fik r s- 3— 5> pl- »i-> figs. 9— IO )- 



* Smith erects for this the genus Calosatumia, see Proc. U.S. Nat. Mils., ix., p. 

 432 (1886). 



t We are inclined to disagree entirely with this. Our own impression is that 

 this repi'csents extreme specialisation. 



