152 Mr. Poulton's notes in 1885 upon 



fertile ova must be exceedingly difficult to get, and, I 

 believe, have never yet been obtained. 



In the hope of being able to examine the earlier stages 

 of this larva at some future time, I will now give a brief 

 description of what I believe will be the main results of 

 such an investigation. In the early stages the caudal 

 horn will possess a forked apex, with a bristle forming 

 the termination of each prong. It is probable that the 

 horn may be movable like that of S. ligustri. The larva 

 in the first stage will prove to be hairy, and will almost 

 certainly possess two dorsal and two lateral rows of 

 longer hairs, and the caudal horn will bear both kinds of 

 hairs ; later the larva will be shagreened, the dots being 

 the tubercles at the bases of the hairs, and in all 

 probability the rows of longer hairs will give rise to larger 

 dots, which will have a special relation to the markings 

 as in the early stages of S. ligustri and Siiierintlius 

 larvae (see Trans. Ent. Soc. of Lond., Part II., August, 

 1885, p. 281). The larva will be shagreened until the 

 end of the penultimate stage (probably the fourth), and 

 probably the special modification of the dots in the 

 dorsal region only takes place in the last stage. The 

 first markings are probably like those of the young 

 S. ligustri, — a subdorsal and a line parallel to and above 

 the latter upon the thoracic segments, seven oblique 

 stripes, and an " eighth stripe." All these markings 

 will be light-coloured (probably white, like those of 

 S ligustri) ; later the subdorsal will fade, and the stripes 

 become prominent (the reverse conditions obtaining 

 primarily), but the anterior part of the subdorsal may 

 be traceable in the third and perhaps the beginning of 

 the fourth or even fifth stages. Since writing this 

 I have detected traces of the subdorsal in an adult 

 brown variety (see above). It is probable that the 

 extreme prominence of the coloured borders and the 

 fading away of the stripes is a very late feature (perhaps 

 only in the last stage), while the borders were at first 

 represented by darkened ground colour. I am inclined 

 to think that the borders arose in the normal way, as a 

 modification of this darkened ground colour probably in 

 the third stage, and that the accessory aid afforded them 

 by the presence of coloured patches developed from the 

 bases of shagreen dots is a late feature, and has nothing 

 to do with their origin. But at the same time the final 



