the Egg-laying of Trichiosoma. 179 



passage the egg is drawn out into a long spindle shape. 

 The egg gradually appears at the anterior margin of the 

 saw, the saw working as during cutting, not at its base but 

 along its whole length, and whilst the saw is still in the 

 position in which it stopped cutting. Gradually the egg 

 grows larger and, in doing so, the saw passes back into the 

 lower portion of the space cut, apparently pushed by the 

 increasing egg, and when the saw is quite pushed back the 

 whole egg is seen to be present and the saw (terebra) is 

 in a second or two withdrawn. At the first appearance 

 of the egg only a narrow margin is occupied by the portion 

 present, and one can only suppose that the shell is suffi- 

 ciently extensible for the rest of it to be in the basal portion 

 of the saw and the body of the insect." When half laid 

 it must have quite a dumbbell shape, one rounded end in 

 the pocket, the other within the body of the insect, the 

 narrow co/mecting portion in the neck of the terebra. 



It may be observed here that in Phymatocera the terebra 

 was withdrawn after the pocket was cut and replaced to 

 lay the egg, but in Trichiosoma no such break occurred, 

 only a short rest being taken with the terebra still in situ. 



Mr. Morice, indeed, relates that the process of rotation 

 towards the body of the insect, apparently precisely Uke 

 that in Trichiosoma, was not as in Trichiosotna stayed when 

 a pocket was formed, but continued until the terebra cut 

 its way out, and in the result, the process being repeated, 

 there was a long clean cut or slit wath a succession of eggs, 

 laid along it. In Trichiosoma, the original incision of 

 entry is never enlarged, and each egg is laid in a separate 

 and independent pocket. 



" On one occasion the saw came to a vein in the leaf, 

 which it seemed to find to.be an impediment and it was at 

 once withdrawn. On another occasion the pocket was all 

 but finished, but the original line of entry seemed to have 

 been too oblique, with the result that the forward movement 

 of the saw brought it to the margin of the leaf before the 

 pocket was quite large enough, though to my eye it was 

 very close to full size. This was not, however, satisfactory, 

 and the saw was withdrawn without any egg being laid." 



" April 27th. Sawfly quite refused to experiment wath 

 a slightly rough hairy-leaved willow. Saw two eggs laid 

 to-day on smooth willow. Timed one of these. From the 

 fly settling down to the saw actually entering forty seconds 

 (40"), from this to completion of the straight entry, 



