CH. IX. J THE GALL-mSECT. 163 



not only " making a gfallant show, and being sot up 

 in houses for the decking of the same," but also as 

 " yielding a most cooling aire in the heat of summer." 



" This willow, however," observes Kirby, " is no- 

 thing more than one of the common species, whose 

 twigs in consequence of the deposition of the eggs 

 of a cynips in their summit, there shoot out into 

 numerous leaves totally differing in shape from the 

 other leaves of the tree, and arranged not much 

 unlike the flower of a rose, adhering to the stem even 

 after the others fall off." 



In consistence, these excrescences have nothing 

 in common with the plant to which they are at- 

 tached. On the oak, some are found literally as 

 hard as iron, so as to turn the edge of a knife, while 

 others are as juicy and pulpy as fruit. It is not 

 the tree but the insect which regulates this ; for it is 

 certain that, on the veiy same leaf, one species of 

 gall-fly will invariably form a woody and hard gall- 

 nut, while another as invariably produces a spongy 

 and soft one — although both of these are formed 

 from materials of the same texture. Persons who 

 formerly saw insects emerging from little excres- 

 cences having no visible inlet, were induced to be- 

 lieve that they had either been sucked up by the roots 

 with the juices, or generated by putrefaction. 



Attentive observation, and the use of a lens of suf- 

 ficient power, will explain the manner in which the 

 gall-nut rises. The little fly may be seen settling on 

 the part to which its instinct invariably leads it ; and 

 introducing a sort of sting, its ovipositor, under the 

 epidermis or skin of the plant: it then moves it 

 about as if to enlarge the orifice, and deposites the 

 egg. These eggs, when examined first in the body 

 of the fly, and afterward in the nut, are found to 

 differ so much in size, that Reaumur supposes them 

 to grow after they have been laid. If this be the fact, 

 it is singular, as in that case the egg will not resem- 

 Dle so much that of oviparous animals as that of 



