NOTES ON THE EARLY STAGES OF CEMIOSTOMA LABUENELLA. 187 



the larvae of C. laburnella do not appear to have this power. After 

 having taken larvae out of their mines, I could never induce one to 

 enter another leaf. 



The leaves of the laburnum are composed of three leaflets, and each 

 leaflet is usually large enough to support two or even three larvae of 

 €. laburnella, but when this species is very abundant, many of the 

 leaflets have so many eggs on them that there is little chance of any 

 larvfB maturing. On September 3rd, 1908, I found at Cookham one 

 leaflet with 22, another with 38, and a third with 40 mines. I should 

 ■say that in these cases none of the larvae would reach the fourth 

 stadium, they would all previously die of starvation. Thus, when this 

 species becomes excessively numerous, the females are prone to lay 

 too many eggs on the favoured leaflets, and so, by causing the death of 

 many larvife, to assist in maintaining the proper balance of 

 individuals. 



Most authors place the genera L'ennoHtonia and Phyllootiiitis in very 

 close proximity. However near to each other their ancestors may 

 have been in remote ages, they have become now, if my judgment be 

 correct, widely separated. They are both highly developed genera, 

 but they have developed on diflerent lines. I think the feature of 

 greatest contrast is shown in the pupa. In C'eniiostoma the pupa may 

 perhaps be said to be armed with transverse ridges, and probably the 

 ridges help to retain the pupa in the cocoon. In P/njUocnistiH the 

 pupa is of quite a different form, and is most elaborately armed with 

 spines and bristles to aid it to force its way out of the cocoon. 

 Coniostoma may be near Lyonetia, but if we retain it in the family 

 Lyonetiidae, we must remove Fhyllocnistis, which, in my opinion, 

 rather belongs to the Gracilarvidae. When I first began to observe 

 CeuiioHtoma laburnella, I looked for points of relationship with 

 Phyllocnistis, as most of the authorities place these two genera close 

 together. However, I soon found that my quest met with no 

 encouragement, and that the species I had under observation had 

 apparently no kinship with the insect which so delighted me some 

 years ag3. 



Ovum : — The egg is of peculiar shape, somewhat resembling a pie-dish with 

 the pie in it. This seems to be a shape common to tlie genus, at least the ova of 

 C. spartifoliella and C. lotella agree with C. laburnella in this respect. In profile 

 the outline of the egg rises slightly above the periphery, while below that area it 

 slopes inwardly down to the base. The long axis measures 0"25mm., the shorter 

 0-2mm., and the height 0-06mm. The sculpture consists of slight irregular 

 pentagonal network, which appears on the upper surface only. The colour is 

 milky grey. 



Larva: — First instar. Length 0-6mm., width of head O'lmm. The head is 

 large but the jaws are not prominent. The prothorax is twice the width of the 

 head, the mesothorax rather narrower than the prothorax, and the metathorax very 

 narrow. The body is fairly cylindrical, but the ninth and tenth segments are 

 abruptly narrower. There are no legs or prolegs and no visible tubercles, even 

 under an eighth objective. The usual head marks are pale brown, and the body is 

 pale grey with a yellow central band running down the abdomen. 



Larva : — Second inatar. Length l-2mm., width of head O'lTnim. Somewhat 

 like the larva in first instar but the thorax not so much swollen in proportion. 

 There are traces of the ball-like pads below the meso- and metathorax. On the 

 •dorsum of the prothorax there is *a somewhat quadrangular dark brown patch. 

 The larva mines dorsum uppermost. 



lik-RXK-. — lliird instar. Length *?-8mra., width of head 0-3mm. Head and 

 thorax rather flattened, body more cylindrical. Segmental divisions very well 



