Trigonaspis crustalis. 83 



wards. I allowed the flies from June 6 to 12 to 

 prick indifferently two little oaks, one of which I then 

 kept indoors, and the other out. The next two months 

 passed without any trace of gall formation being dis- 

 cernible. At the end of August I examined some of the 

 leaves, and found eggs in the scars of the punctures 

 which were still visible ; these contained living and 

 moving embryos. At last, on September 6, there 

 appeared simultaneously out of several leaf-veins 

 little whitish galls which grew very slowly, and three 

 weeks more elapsed before they could be positively 

 recognized as those of Biorhiza renum. On one oak 

 sixty, and on the other about seventy galls were formed. 

 The observation of the generation-cycle was thu3 

 complete. 



On account of the morphological features of interest 

 which these two generations offer, I have given a draw- 

 ing of both flies. (Fig. 18-18 ^) 



A comparison of the two generations shows some very 

 striking differences : the form, size, and colouring are 

 utterly unlike, and the variation extends to other parts 

 of the body. In Trigonaspis crustalis the antennae are 

 respectively 14 and 15-jointed, the maxillary palpi 

 5-jointed, and the labial palpi 3-jointed. On the other 

 hand, in Biorhiza renum the antennae are 13-jointed, the 

 maxillary palpi 4-jointed and the labial palpi 2-jointed. 

 Lastly the ovipositor is of a totally different construction 

 (see illustration, PI. III. 6, 6*). Owing to these and 

 other important differences, the two generations would 

 undoubtedly have been considered as belonging to 

 different genera. 



G 2 



