30 Mr. E. B. Poulton's notes upon the colours 



varieties approaching fig. 2 are not uncommon. The 

 alternate arrangement is not marked to the same extent 

 as in fig. 1, and the spots are not greatly developed 

 upon the claspers. Anteriorly the spots are not dis- 

 tributed regularly upon the segments, since the former 

 exceed the latter in number. This fact brings out very 

 clearly the want of relation to the oblique stripes, for 

 these are not developed at all anteriorly, and the rem- 

 nant of the subdorsal, which follows the oblique line 

 system (as will be shown) in the chief protective attitude, 

 sweeps over the contracted anterior segments as a whole 

 without any suggestion of a separate development upon 

 each of them. In fact, a development of oblique lines 

 on each small anterior segment, approaching the de- 

 velopment of the spots, would destroy the symmetry and 

 protective value of the whole system. In fig. 2 there is 

 no tendency towards a drawing out of the spots into 

 lines, except perhaps to some extent in connection with 

 the 7th stripe, as was also shown in fig. 1 ; and this 

 slight tendency does not appear to be in the direction of 

 forming anterior coloured borders. It seems possible 

 that in S. tilim there is a further intensification of the 

 tendency shown in the 7th stripe of ^S*. ocellatus and 

 popuii, and that there is no relation to coloured borders 

 such as are met with in Sphinx. 



If >S'. tilicB is developing after the manner suggested by 

 Prof. Weismann, and Sphinx ligustri has lost the spot- 

 stage in its ontogeny, it must nevertheless remain true 

 that S. ocellatus and popuii have progressed in quite 

 another direction. But, considering the above facts and 

 those which follow, it seems more probable that S. tilice 

 occasionally presents us with a modification which 

 renders its own oblique striping more distinct by a 

 suffusion with the colour of the spots, but that this has 

 no significance for the coloured borders invariably 

 present in Sphinx ligustri, &c., which arise directly from 

 the darkened ground colour anterior to the light stripes. 

 Such an extreme instance of this variety as is shown in 

 fig. 2, coexisting with a vast majority of larvae (of the 

 same species) without a trace of the marking, seems to 

 be greatly in favour of the view that the whole character 

 is due to occasional reversion to a form of marking 

 which is disadvantageous to larvae protected by their 

 resemblance to leaves. The theory of protection by 

 resemblance to galls is not supported by such a complete 



