of certain lepidopterous larvce and pupce. 45 



cannot be doubted that the object of this striking effect 

 and peculiarly prominent method of holding the claspers 

 is to attract attention to the convex side of the bend, and 

 to convey the notion of angularity. After this bend 

 there is a very marked hump on the dorsal side of the 

 12th segment which suggests another bend, only repre- 

 sented, in fact, by a very slight downward curve ; while 

 the last pair of claspers aids the delusion by being 

 directed downwards and backwards. (It may be note- 

 worthy that these claspers still retain some slight func- 

 tional value in this species). This last curve is also 

 intensified to the observer in another way. After con- 

 templating a series of curves the mind is ready to 

 continue the series upon a very slight suggestion. This 

 principle is frequently made use of in the markings of 

 insects, and I have had occasion to allude to it in 

 describing the markings of the genus Smerinthus. Thus 

 in N. ziczac the eye is directed to the convex sides of all 

 the curves, and the impression is conveyed that they are 

 sharp and angular, that they follow the outline of the 

 structures on the convex sides instead of the true outline 

 of the larva. All the four curves are, in fact, quite 

 smooth and gentle, and the last one hardly exists. 



When the larva is crawling it is seen that very much 

 of the effect is due to the prominent way in which the 

 structures are held, for they are then hardly noticeable 

 (except the humps on the 6th and 12th segments, and 

 the former appears now on much more equal terms with 

 that on the 7tli segment). 



On October 15th of the presentyear (1883) I found a full- 

 fed larva of iV. ziczac (feeding on Salix rubra near Oxford) 

 which presented some differences from the normal form. 

 It was much darker than usual ; the spiracular line — 

 really a little below the spiracles — was distinct on the 

 2nd and 3rd segments, then absent, reappearing faintly 

 on the 6th segment, and increasing in distinctness on 

 succeeding segments to the 10th. On the clasper of the 

 10th segment it formed a very prominent bright yellow 

 stripe, bordered posteriorly by the dark colour of the 

 posterior segments, here also darker than elsewhere. 

 The spiracular line was completely turned aside into the 

 clasper, and there was no trace of it posteriorly to this 

 point. Dorsally the dark longitudinal patch on the 2nd, 

 3rd, and 4th segments is seen to be part of a regular 

 dorsal line traceable up to and over the hump on the 

 12th segment, even down to the anal flap. The dorsal 



