^r. ]l. Taylor 



61 



35 to 50 seconds, the duration being influenced no doubt by the species 

 and age of the leaf, the vigour of the fly and other factors. In captivity, 

 a fly sometimes oviposits in a leaf while crawling on the upper surface, 

 and on these occasions, probably because the tissue is more compact on 

 this side, the process lasts somewhat longer. For example, it was noticed 

 that a fly which spent about 50 seconds in passing an egg into a piece 

 of celery leaf from above, required about 10 seconds less when inserting 

 an egg from below. 



At hatching, which takes place in about six days, the larva breaks 

 through the egg-shell at the end pointing away from the puncture in the 

 cuticle, and burrowing forwards, eats out at first a narrow gallery in the 

 tissue, but after a few days changes its mode of attack, and mining in 

 all directions gives rise to a wider cavity less regular in shape. Increasing 

 in size as the larva feeds, the cavity thus produced often coalesces with 

 others of the same kind, and, with these additions, forms a compound 

 blister containing it may be several larvae — a circumstance which has 

 perhaps given rise to the existing impression that the Celery Fly, like 

 Pegomyia and other blister-making flies, lays its eggs upon the surface, 

 whereas it resembles more closely such flies as the Phytomyzidae, which 

 lay their eggs in the interior of leaves. 



Celery-fly Egg in leaf of Hcracleuni. showing puncture made by fly. 

 (Drawn by Miss E. M. Wright.) 



