4 Earlier Views. 



a large number of larvae, and in June there emerged 

 flies of each sex in nearly equal proportion. Late in 

 the summer there was formed on the points of the 

 young shoots of the same oak tree, a differently shaped 

 gall in which the flies passed the winter. This latter 

 species occurred in the female sex only. It closely 

 resembled the former species, but was somewhat larger. 

 From this observation Bassett arrived at the conclusion 

 that each species of Cynips which is found occurring 

 exclusively in the female sex is succeeded by another 

 generation which is bisexual : and this he contended 

 was entirely opposed to Walsh's hypothesis. Bassett 

 concluded by saying that it would not surprise him 

 if it were proved that every species of the genus 

 Cynips had each year two generations differing in the 

 manner indicated by him. Had I been acquainted 

 with Bassett's work in the year 1875, when I began to 

 investigate the Cynipidae more carefully, I should 

 probably more easily have found the key to the 

 mysterious problem of their propagation. A lucky 

 chance led me to select the species of Neuroterus for 

 my first experiments : these galls are easily collected in 

 large numbers, and there is little difficulty in rearing 

 the flies. In every case I made a point of breeding 

 the flies from the galls, so that I might be absolutely 

 certain of the species. The flies emerge in March and 

 April from Neuroterus galls which had matured in 

 autumn, and they proceed at once to lay their eggs in 

 the buds of the oak. It struck me as remarkable that 

 although the egg was laid so early, the gall did not 

 develop until July : and it was the strangeness of this 



