Apliilob'ix collaris. 45 



lays its ^%% in an immature gall, the growth of the gall 

 is altered with the death of the original larva, and 

 becomes pathological. Sometimes these morbid galls 

 remain small, at others they become firmly rooted into 

 the parent tissue. (Fig. 9.) 



Rearing the fly. The same precautionary measures 

 must be taken as in the former species of Aphilott'ix 

 in order to obtain the flies from the galls. The duration 

 of the larval state is the same. After the galls are 

 mature, a year and a half passes before the appearance 

 of the flies. 



Fly. 3 mm. long. Head and thorax dark, often with 

 reddish lines on the back ; thorax smooth and shining ; 

 scutellum reddish brown, dull, and hairy ; abdomen dark, 

 base sometimes reddish ; legs orange ; coxae always 

 dark, and sometimes also the bases of the femora. 



Experimental breeding. This fly has hitherto passed 

 as rare, chiefly because the galls are difficult to find, and 

 the rearing does not always succeed. Further observa- 

 tion has shown me, however, that in some years it is 

 very common. I made my first attempts at breeding in 

 1876 with two flies, which pricked several buds between 

 April 4 and April 6. The eggs were always laid in 

 the rudimentary leaves in the centre of the bud, and 

 accordingly it proved, as might have been expected, 

 that the galls developed on the leaves. As soon as the 

 buds had unfolded, a tumid thickening was observed on 

 two of the leaves. This was the beginning of a gall 

 which was soon recognizable as Andricus ciirvator. 

 I repeated the experiment in 1878, putting six flies 

 on a little oak, the buds of which they continued to 



