104 



BRITISH LEPIDOPTERA. 



not in a rounded tubercle but, in a sharp spine curved backwards 

 to form a hook. The labral prominence is less marked. The 

 dimensions are : 





DISTANCE 





ANTERO- 





FROM FRONT 



TRANSVERSE 





MEASUREMENTS AT 







POSTERIOR 





OF 



DIAMETEP. AT 







PUPA. 





DIAMETER AT 



Eye-spines 



i-omm. 



4'3mm. 



4'Oram. 



Posterior margin of prothorax 



4'° » 



5'° >> 



6-o „ 



Middle mesothorax . . . . . . . . 



9*° » 



7'0 „ 



7*o » 



End of 1st leg 



10-5 „ 



7-4 »< 



7'8 „ 



End of 2nd leg 



12-0 ,, 



7'8 ,, 



8-o „ 



Widest part 4th abdominal 



16-5 ,, 



8-o „ 



8-6 „ 



End of 4th abdominal 



18-0 „ 



8-o „ 



8-2 „ 



Spines 5th abdominal 



21'0 ,, 



85 >, 



8-o „ 



Spines 6th abdominal 



23-5 » 



7"8 „ 



7*5 >> 



Spines 7th abdominal . . . . 



26-0 ,, 



7-o ., 



6-o „ 



At 8th abdominal 



28-4 ,, 



5'0 „ 



4*5 » 



At base of spine 



3 1 ' » 



2-6 ,, 



1-2 „ 



Total length 



33'4 1, 







The rows of spines on abdominal segments 5, 6, 7, differ by passing dor- 

 sally more to the front of the segment, and breaking up there into more 

 numerous smaller spines. Laterally, the ridge passes straight round 

 the pupa, at some distance behind the spiracles. In E. elpenor it 

 seems as if it would like to go through the spiracle, and bends a 

 little in order to pass round behind it. The line is, indeed, broken 

 by the spiracle, in some specimens absolutely, but, in most, a few 

 fine spines exist behind the spiracle. In T. porcellus, except a little 

 bend in the ridge on 7th segment, the spiracle does not enter into 

 the question. The anal spine seems to be the same in both, unless 

 in T. porcellus the dorsum be more rounded, as one surface, instead 

 of almost divided into an anterior flat and posterior sloping surface 

 as in E. elpe?wr. The wings are much more distinctly and neatly 

 marked than in E. elpenor. Poulton's line is crisply marked by special 

 wrinkling beyond it. In E. elpenor the surface beyond is smoother. 

 The nervures are often (there is a good deal of variation) well marked 

 by broad pale lines between darker spaces and with rows of dark 

 dots down them, nervure 8 being almost lost in the costa. There 

 is no scar of caudal horn, except, in one specimen, a depression, 

 which may be accidental. The absence of horn scar is notable, as it 

 exists both in Pterogon proserpina and Thaumus vespertiliOy the only pupae 

 I have of Sphinges that I know to have hornless larvae. The sculp- 

 turing is wrinkling throughout; there is no pitting ; on the 8th abdominal 

 there is well-marked transverse wrinkling (cerebral pattern) and on 

 9th there is fine longitudinal wrinkling (dorsally). The pitting of 

 anal spike is much as in E. elpenor. The subsegmentation 

 seems to be identical. When we examine the pupa-skin micro- 

 scopically, the differences between T. porcellus and E. elpenor take 

 at first view a different aspect, in that porcellus seems to the naked eye 

 to be wrinkled throughout, E. elpenor to be only or chiefly pitted, 

 on 8th, 9th, and 10th abdominal segments. Microscopically, the dark 

 circles which we may call pits or hair-points are much more numerous 

 on these segments in T. porcellus than in E. elpenor. In E. elpenor 



