102 Production of Wimjed Forms in an Aphid 



it was desirable to obtain a high proportion of these forms in each genera- 

 tion; therefore, I repeated Shinji's experiment. The numbers used were 

 necessarily small, but since my results are not altogether in agreement 

 with his, and as the subject is of some biological and economic interest, 

 they are given here. 



The aphid used is a common pest of red currant bushes, and, during 

 the summer, also feeds facultatively on weeds such as Lamium and 

 Galeopsis. An account of its life-cycle has been published elsewhere (4). 



Cuttings of red currant were planted in nurseryman's sand previously 

 washed for 24 hours in tap- water, which, according to Shinji, is one of 

 the "non-wing producing" agents. A single stem mother or fundatrix 

 was placed on each, and the young born were removed within 12 hours 

 to similar cuttings watered either with tap-water, or else with an m/50 

 solution of MgS04, the latter being about the optimum strength of this 

 salt for wing production, according to Shinji's observations. 



Table I. 



The results of the experiments are given in Table I. They were not 

 carried further than the second generation since the object of the work 

 was not primarily to repeat Shinji's experiments; and it was found that 

 the proportion of winged forms in each strain was high in the first two 

 generations, whether the salt solution was used or not. It will be seen 

 that where MgS04 was employed for the first generation, the percentage 

 of winged individuals varies from 70 per cent, to 4 per cent, and that the 

 highest proportion of all was obtained in strain J, where nothing but tap- 

 water was used. 



A second experiment was made in May. It was necessary, to ensure 

 the continuance of certain strains, to transfer the third or fourth genera- 



