and reproduction rates are affected more by the quality than by the quantity 

 of the nutrients which the aphid gets from the plant and are, therefore, 

 dependent on its feeding site. When aphids live on leaves Avhere their food 

 may be deficient in some important nutrients, or when they are undernourished, 

 they may benefit from an increase food suj^ply caused by ant-attendance an 

 so they may grow and reproduce faster, as indicated by El-Ziady. If the 

 aphid is to benefit nutritionally from the stimulation of feeding, it must be 

 able to assimilate more of its food; because it apparently does not do so, 

 it seems that only the ant benefits nutritionally from the association. 



A more important effect of ants on the aphids they attend seems to be the 

 inhibition of the production of winged forms first reported by El-Ziady and 

 Kenned}^ (for A. fabae attended by L. niger), but recently reported for 

 A. craccivora by Johnson in those experiments first and second instar nymphs, 

 which would normally have become allatae on mature bean leaves in water, 

 develojDed into apterae when attended by the ant, Parafrachina baveri Mayr. ; 

 third and fourth instars were not so affected but continued to develop as 

 allatae. Johnson thought that although the stimulation of feeding may be 

 involved in the aphid's change of form, it is more probable that endocrine 

 activity changes as a result of being attended by ants, and this might happen 

 independently of nutrition. 



Ant-attended and ant-free aphids do not differ in size or in rate of growth, 

 but this is not evidence that nutrition is unconnected with change of form; 

 and a change in endocrine activity might itself depend on the quality of the 

 food. We know that for aphids the nutritional value of leaves of different 

 ages varies, that alate aphids are more often produced on old leaves and plants 

 and apterae mainly on young vigorous growth. If old leaves or plants contain 

 some substance or substances in the sap which inhibit wing formation (direct - 

 \y, or though an endocrine change or other indirect means), an increase in their 

 supply to an ant-attended aphid might divert it from the potentially alato 

 to the apterous condition. 



It is interesting to note that Johnson has brought about the change from 

 the potentially alate to the apterous state in A. craccivora by two methods: 

 by parasitization with Ajjhidius (1959 a) and by ant-attendance (1959, b). 

 The reverse change (potentially apterous to alate) appears not to have been 

 reported and perhaps can not be made. I could not confirm the delay in pro- 

 duction of alatae in ant-attended A. fabae colonies reported by El-Ziady and 

 Kennedy (Banks, 1958). The aphids used were the progeny of alatae which 

 are nearly always apterous so that Avhen they were attended by ants they 

 could not perhaps be changed to the other form, ant this may explain the dif- 

 ference between my results and those of El-Ziady and Kennedy. 



Wing formation in aphids appears to be stimulated by such means as over- 

 crowding, parasitization and perhaps intermittent starvation, or even handling 



331 



