74 



The young ]arvt>? differ from the mature ones l\v being- more con- 

 tracted, thus appearing more strongly tuberculate at the sides. More- 

 over, very 3'oung hirvte are entirely whitish. 



The pupa (fig. 3, c) is at first whitish, but soon assumes a nnifonn rich honey color. 

 In general shape it resembles the full-grown larva, though a little shorter and some- 

 what more convex. The head has become more convex and especially wider in 

 comparison with the first thoracic segment; the second and third thoracic segments 

 have of course changed the appearance of the wing pads, which are longitudinally 

 striate and bent downward along the sides of the body, resting between the second 

 and third pairs of legs and ref ching with their extremities to the posterior end of 

 the first ventral segment. The abdomen still shows the lateral tubercles so con- 

 spicuous in the larva and these tubercles are furnished each with two or three 

 stiff bristles. The edges of the median transverse impressions on the dorsal sur- 

 face of the abdominal joints of the larva are in the pupa much sharper, ridge-like, 

 and also furnished with sparse bristles. The fourth abdominal spiracle is replaced 

 by a long stout spine directed backward. On the ventral surface the armature of spines 

 is even more conspicuous than on the dorsal, each segment having a sharply raised 

 transverse line which is sinuate and more raised at the middle and there furnished 

 with five or six setiferous tubercles. On the penultimate segment the transverse 

 line is not sinuated and is furnished with eight tubercles, the two median ones being 

 much less prominent than the lateral ones. The anal segment is also furnished with 

 two obsolete transverse lines and with a few inconspicuous tubercles. 



This array of l)ristles and tubercles enal)les the pupa to move rapidly 

 forward or backwaixl by a wriggling motion of the abdomen. The 

 power of locomotion exhibited in this pupa is really astonishing; in 

 fact, the pupa is nuich better afoot, so to speak, than the larva." 



The mine produced by the larva has not hitherto been described so 

 that it ma}" be distinguished from the mines produced by the various 

 leaf-mining Tineidie which are usually w^orking in company with it. 

 As already .stated, the beetle larva consumes the whole of the paren- 

 chyma within its mine, thus causing the mine to be equally visible on 

 both sides of the leaf, whereas in the case of the Tineidw the larva 

 destroys only a thin layer of the upper portion of the parenchyma, 

 causing the mine to be invisible on the underside of the leaf. The 

 Tineid larva very neatly separates the epidermis from the parenchyma 

 so that not a particle of this last is left adhering to the skin-like 

 epidermis forming the roof of the mine, and this roof is of a uni- 

 form pale l)uff' color strongly contrasting with the green of the leaf. 

 On the other hand, the larvie of the beetle accomplish their work much 

 less carefully; numerous small particles of the parenchyma are left 

 adhering to the epidermis, and consequently the latter forms a much 

 thicker covering to the mine than in the case of the Tineidw, The 

 color of the mine is a pale green slightl}^ tinged with brown, its surface 



« The power of locomotion in the pupa becomes still more remarkable if we take 

 into consideration that the pupa, which always remains within the mine, has really no 

 occasion or opportunity to make use of it, unless we except the possibility that by 

 wriggling about in their mines the pupa3 might succeed in evading parasitic or pre- 

 daceous attack. 



