Maud D. Haviland 10^3 



tion to a second host-plant, and the common Dead-nettle {Lamium 

 purpureum) was selected for this purpose. Winged females of four 

 different strains were transferred to potted plants of Lamium, watered 

 with a m/20 solution of MgS04. Their young were removed to currant 

 cuttings, which, with the exception of one, accidentally left among a 

 number of other plants watered with HgO, were irrigated with the same 

 solution. The numbers reared in this second experiment (Table II) were 

 very small, as the mortality after transference from Lamium back to 

 currant is high, but the only winged forms produced appeared on the 

 plant watered with HgO. This was probably due to chance, but it seems 



Table II. 



to show that the presence or absence of magnesium is not the only 

 determining factor in wing production. It is important to have controls 

 in experiments of this kind, for the appearance of a large number of 

 winged forms in the first generation (Table I) might have been attributed 

 to the salt solution used to irrigate the plant, if other strains, reared 

 under identical conditions but without the magnesium sulphate, had 

 • not likewise produced a high proportion of these forms. Moreover, the 

 proportion of winged forms among the magnesium treated broods varies 

 considerably. The evidence points to the conclusion that, as regards this 

 species at any rate, the appearance of alate females cannot be attributed 

 wholly to the presence of magnesium salts, although it is quite possible 

 that the aphides may react to metabolic changes in the host plant, in- 

 duced by an abundance of such salts in the soil. 



Gregory (3) found that the proportion of winged to wingless females 

 in the broods of the pea aphid, Microsiphum destructor, could be raised 

 by periodically starving the parent during its development. This lends 

 some support to the view that exhaustion of the host-plant leads to the 

 production of winged migrants which will seek fresh food supplies for 

 the maintenance of the race; but it may be pointed out that periods of 

 total abstinence, alternating with periods of normal feeding, would not 

 necessarily produce the same effect as continuous feeding upon an in- 

 adequate diet. 



