of certain lepidopteroiis larrce and pupa:. 37 



the white lines, and in the presence of borders formed 

 by a deepening of the ground colour. The pure white 

 stripes only appear when the skin becomes smooth, but 

 before this the purple borders can be seen, although 

 narrow and not conspicuous. I do not remember any 

 indications of spots which gave rise to the borders by 

 coalescence, but I believe that the latter took the place 

 of the dark green edges. 



The above statements are from memory : I can now 

 speak with certainty as to the last stage, for I have care- 

 fully examined several specimens after reading Weis- 

 mann's book. The beautiful delicate white stripes still 

 retain indications of the shagreen covering in their 

 anterior inferior extremities, which are made up of 

 scattered white points. Superiorly and posteriorly they 

 still show traces of their origin from stripes, resembling 

 those of Smerinthns, for, continued on to the segment 

 behind that in which most of its course lies, the white 

 stripe becomes of a pale yellowish green, and can be 

 followed nearly up to the dorsal line (formed by the 

 dorsal vessel). The recognition of this superior con- 

 tinuation needs close observation, but it will be readily 

 seen now that attention is directed to it. At the same 

 limit the purple border to the white stripe changes 

 superiorly and posteriorly into a dark green border to 

 the yellowish stripe. This also continues nearly up to 

 the dorsal line, and can be recognised with a little atten- 

 tion (see fig. 7). These observations seem to prove that 

 the purple border has been modified from a dark green 

 border (like that of Smerinthns) at the sides, but not 

 above, where the latter faintly persists. In the same 

 manner the pure white stripes have arisen from lines, like 

 those of Smerinthns, which still remain above. The relation 

 between the bright colouring of Sphinx and the obscure 

 colouring of Smennthns is very well seen by looking care- 

 fully at an adult Sjihinx lifiustri from above (see fig. 6). 



These last facts, together with the marking of earlier 

 stages and the long retention of shagreen, prove that the 

 beautiful colours of Sjihinx ligiistri have been acquired 

 very late in the phylogeny. 



8. Further notes on the adult larva of Sphinx ligiistri. 

 — The anterior spiracle (on the 2nd segment) is _ not 

 hidden during rest, although the Sphinx attitude is so 

 marked in this genus. I found one specimen of S. 

 ligiistri (at Wootton, near Oxford, Sept. 14th, 1883) 



