10 [January, 



exaiiiining these petioles it seemed to be clear that the fly kid her eggs 

 on either side of the leaf-stalk, but always from the upper margin, but 

 on the 29th, a fly having been induced to lay on a fresh petiole, she was 

 seen to do so from the under margin ; from this, and further obser- 

 vations, it appears that the margin selected is a matter of accident or 

 convenience, and that the margin that happens to be uppermost is most 

 likely to be used. The petioles of the leaves of this poplar are flattened 

 from side to side, especially towards the leaf where the}" are very deep, 

 from upper to lower margins, but narrower than nearer the stem. They 

 are therefore rather flat on each side, so that the laying is under a flat, 

 rather than a curved, surface. The middle of the petiole is the selected 

 portion, the more rounded base and very thin end being avoided. The 

 two sides are indifferently occupied. The six leaves above referred to 

 presented the following as the numbers of eggs laid on each side : — 

 8-11, 10-10, 9-9, 12-24, 7-0, 8-5. The eggs vary a little in their 

 spacing, each fills a sepai*ate pocket, each pocket is about 1*5 mm. 

 (measured along petiole), and the space between pockets may be from 

 0-7 to 1-0 mm. 



In the process of la34ng the cuticle is pierced at the upper (or lower) 

 margin of the side of the petiole by the sharp tip of the saw. The saw 

 advances very rapidl}'', keeps all the time close under the cuticle, going 

 fii-st straight down and then cmwing up, just as in Trichiosoma, or 

 Fhymatocera, till it is directed forwards from the opening of entry, 

 and then the e^g is laid in the same apparently magic way as in those 

 genera. During the laying the terebra retreats to the position of first 

 complete entry and is then withdrawn. The insect then advances a step 

 to the position for the next pocket — the whole process is miuch m.ore 

 rapid than in the other species referred to. 



The saAvs seem to be always quite straight and have an extremity 

 narrowing to a sharp point. This dagger-like point makes the first entry 

 and continues the first straight advance, but the widening to take the 

 whole width of the terebra is done by the margin of the saws, and 

 practically the whole of the cutting, as the terebra sweeps round to form 

 the cavity of the pocket, is done by the margins of the saws. In pene- 

 trating, the points of the saws advance and retreat in relation to the 

 supports, and the points of each saw advance alternately ; this alternate 

 advance seems to have less amplitude than their combined advance and 

 retreat. It may be described better, perhaps, by saying that the two 

 saws advance together and then retreat, but at one advance one saw is in 

 front, in the next the other, and so on alternately. The sharp stiletto of 

 the saws is thrust a long way beyond the support and then withdi'awn, 



