Andncus nodidi. 33 



happens that two or more flies lay their eggs close to 

 each other at the same spot. I cannot conceive how 

 otherwise the colossal compound galls sometimes found 

 could be formed. On one occasion I collected 1,100 

 flies from one Aphilotrix radicis gall ; as, however, 

 a single Andricus nodidi fly only carries about 500 eggs 

 in its ovarium, it is clear that such a gall could only 

 be produced by the co-operation of several flies. 



On the little oak tree pricked in August, 1878, I was 

 able to Vv^atch the subsequent development of the galls. 

 In September a thickening of one of the pricked spots 

 gradually took place, the bark was raised and at last 

 broken through, and a new structure, of a hemispherical 

 form, was obtruded. When, in October, vegetation 

 ceased and the leaves began to fall, the growth of the 

 gall stopped, but only to resume its course next spring. 

 In this first stage of its growth the gall had the con- 

 sistence of a potato, and sections could easily be made 

 for the purpose of microscopical examination. The 

 apparently homogeneous tissue was found to be studded 

 with countless minute larva chambers. In the centre 

 of each of these chambers lay a very small larva 

 surrounded by a succession of concentrically arranged 

 cells. The cells next the larva were the largest, were 

 filled with starch granules, and some were undergoing 

 degeneration. The cells of the more remote layers 

 became steadily smaller until they finally passed into 

 the surrounding cambium tissue. Here and there 

 vascular bundles pushed their way into the newly 

 formed stroma, and brought it thus into closer con- 

 nexion with the parent tissue from which it grew. 



D 



