upon lepidopterous larva, dkc. 285 



appearance of these features, but they also conclusively 

 show that the characters do not necessarily develop in 

 the successive stages, but that they may gradually 

 decrease and disappear in the course of development, 

 just as if they were features of phylogenetic significance 

 only. Extended observation will, I feel sure, confirm 

 this conclusion, and will prove that the spots are 

 present upon a much larger proportion of young larvae 

 than upon those in the later stages, although it is well 

 known that in a certain proportion of the latter the 

 character reaches a pitch of perfection which has not 

 been hitherto described in the earlier stages. The first 

 appearance of the spots upon the third abdominal 

 segment, and then upon the second thoracic, has 

 already been proved by Mr. White, — a conclusion 

 which is abundantly confirmed by my observations. 



On September 21st I found two larvae of this species 

 upon balsam poplar at Bembridge (Isle of Wight). Both 

 possessed the spots to a remarkable extent. One larva 

 was about half-grown in the last stage, and was of a 

 yellowish green ground colour. The other was adult, 

 and a very white variety, exactly resembling in ground 

 colour that figured by Mr. Bignell, and represented in 

 my paper (Trans. Ent. Soc. Lond., Pt. I., April, 1884). 

 The spots of the upper row in both larvae are arranged 

 below in the order of relative magnitude, the largest 

 spots being numbered first : — 



Yellow larva — (1) abdominal, 7 ; (2) abdominal, 3 ; 

 (3) abdominal, 4 ; (4) abdominal, 2 and 5 ; (5) abdo- 

 minal, 1 ; (6) thoracic, 2 ; (7) thoracic, 3 ; (8) thoracic, 

 1, and abdominal, 6. 



White larva — (1) abdominal, 3 ; (2) abdominal, 4 and 

 7 ; (3) abdominal, 5 and 2, and thoracic, 2 ; (4) abdo- 

 minal, 1 ; (5) abdominal, 6, and thoracic, 1 and 3, — all 

 these very small. 



There was also a reddening of the upper side of the 

 base of the horn in both larvae. 



Nos. (1) and (2) of the yellow larva were almost 

 equal, so that the spot which appears first in the young 

 larva is one of the largest in the adult, but the same 

 comparison does not hold with the spots which appear 

 next in order. It is strange that the spots should be so 

 large upon abdominal 7, for in S. ocellatus this segment 

 is remarkable for the minute size of its spot, which is 



