of certain lepidopterous larvce and pupcs, 31 



system of spots as is shown in fig. 2, and it is further 

 rendered unlikely by the fact that S. popidi presents the 

 variety far more commonly and to a far greater extent 

 than S. ocellatiis (considering the relative abundance and 

 arrangement of galls upon the leaves of their respective 

 food-plants). The spotted varieties are certamly more 

 conspicuous than the others, and such a variety as fig. 2 

 is very much more conspicuous. Not only is this true, 

 but the modification of the spots, suggested by Professor 

 Weismann, would also be disadvantageous m rendermg 

 the larva conspicuous, and Sphinx ligustri is far more 

 easily detected than the Smerinthus larvae. The pro- 

 tective resemblances, in fact, belong to two different 

 classes (as will be pointed out), for S. ocellatiis resembles 

 a rolled-up leaf (at any rate when feeding upon apple), 

 while S. ligustri is protected by a general harmony with 

 its surroundings. Hence the coloured borders would be 

 disadvantageous in the former case, as they would 

 destroy the special resemblance; while they may be 

 beneficial in the latter case, where the resemblance is to 

 a general effect caused by complex combinations of light 

 and shade falling upon a large and heterogeneous object. 

 And so also with S. tilics and populi, which are protected 

 by a special resemblance to leaves or parts of leaves. 



2 The origin of the white stripes in Smerinthus. — If the 

 shagreen dots be carefully observed they will be found to 

 possess an annular arrangement. There are generally 

 eight rings in the largest segments. As each ring 

 intersects a white stripe (either belonging to the oblique 

 or subdorsal system), the dot that is placed upon the 

 stripe is much larger than others near it m the same 

 ring The largest oblique stripe— the last -has more 

 than one row of dots. When the full-fed larva has 

 ceased to feed, and is lying quiescent m a contracted 

 state before pupation, the colours are much changed. 

 The white striping disappears, and the ordinary shagreen 

 dots become very inconspicuous, but the enlarged dots 

 remain distinct, and thus the whole system of lines can 

 at once be recognised. It is thus seen how large these 

 dots really are, for under other conditions they are hard 

 to detect, being placed on a white ground. In addition 

 to their arrangement upon the usual stripes there is a 

 line of enlarged dots on each side of the dark pulsating 

 dorsal line formed by the dorsal vessel. In this case 

 the ground colour has not become white along the line 



