THE GALL-FLIES. 315 



is attached to tiie thorax by a very small portion of its dia- 

 meter, and is often lengthened into a more or less slender 

 footstalk. The females are furnished with an ovipositor com- 

 posed of several valves, and similar in most respects to that of 

 the Sires- Mr. Westwood arranges them in two groups ; the 

 one, wliich he calls Spiculifera, having ' two delicate spicula, 

 working in a horny semi-canal, which is defended at rest by 

 two often partially exserted valves.' The second division in- 

 cludes those in which the abdomen is terminated by a tele- 

 scopic retractile tube. 



The first family is the Gynipidce, popularly known as Gall- 

 flies, in which the ovipositor is internal and more or less spiral, 

 and the antennae are straight and have from thirteen to fifteen 

 joints. By means of the ovipositor, the female insect punctures 

 various portions of plants, the ribs and nervures of leaves, 

 young twigs and roots, being the favourite objects, and by the 

 same instrument she introduces an egg into the wound, to- 

 gether with a drop of some irritant liquid. The effect of this 

 liquid is very curious. It mixes in some way with the sap of 

 the tree, which causes r swelling to take place, in the middle 

 of which the egg is hatched into a larva, and finds at once its 

 board and lodging combined. As is the case with tlie Saw- 

 flies, the egg of the Gall-fly enlarges after it is deposited, 

 imtil it is three or four times as large as when it was first de- 

 posited. 



The varieties of galls, in shape, colour, and size, are almost be- 

 yond calculation ; for, despite the enormous niunber of Gall-flies 

 that are already known, new specimens are still being discovered, 

 and every species makes its OAvn gall. How the instilled liquid 

 acts no one knows, and there are few more curious problems in 

 nature than that which the growth of the gall involves. Were 

 each kind of tree, for example, to produce one kind of gall, 

 it would be easy enough to understand that the irritating liquid 

 introduced by the insect would produce a certain sort of ab- 

 normal swelling. But when we find that a variety of Gall-flies 

 attack one tree, such as the oak, and that each produces an 

 entirely different gall, the problem is a very perplexing one. 

 In many cases, such as the well-known currant-gall, the nut- 

 gall, and the hard, woody gall of Gynips Kollari, which is shown 

 on Plate IX. Fig. 6, each gall has but one inliabitant. But in 



