BIOLOGICAL STUDIES OF APHIS RUMICIS, LINN. 87 



appear to be an adaptation of the species to its poly]:)hagous habits. Thus when the 

 w.v. o $ aUght on a new plant a good colony is established on that plant by a.v. $ $ 

 being produced. On the other hand, the liability to overcrowding as the plant 

 becomes heavily infested is overcome, and the wider distribution of the species to 

 other host-plants is ensured by the production of winged forms in due course. 



Towards the end of summer, after a number of agamic generations have been passed 

 through on the intermediate hosts and at a time when suitable intermediate hosts are 

 becoming scarce, there are produced winged viviparous females, which are physio- 

 logically specialised, but morphologically resemble the other alienicolae alatae. 

 These are the sexuparae alatae or remigrantes. They fly back to the winter host, on 

 which they produce true oviparous $ $. 



At about the same period winged (^ (^ are produced from certain of 

 the alienicolae apterae (sexuparae apterae) on the intermediate hosts and fly to the 

 winter host, where the sexual $ $ are fertilised. Fertilised eggs are then laid on 

 the winter host, near the buds, or in crevices in the bark of the older branches. 

 These over-winter and hatch out in spring, producing the fundatrices. 



The alienicolae apterae on the intermediate hosts gradually die out, owing partly 

 to the tendency to produce w.v. $ $ or winged sexuparae and (^ cJ, and 

 partly to unfavourable conditions.* By confining the Aphids to broad bean plants, 

 sexual forms were produced on these plants in due course. f 



In my experiments it was found that the alienicolae alatae of any generation could 

 be transferred back to Euonvmus, on which plant they produced young, and eventually 

 in succeeding generations both sexual (^ (^ and sexual $ $ appeared. Further, it 

 was found that even if the Aphids are bred in successive generations on Eitonymns, 

 sexual ^ (^ and 2 5 will appear in due course. It should be noted, however, that 

 young growth was ensured on the Enonymus bushes by cutting them back. Males 

 were first noted on 10th August 1920. 



Similarly on beans, on which intermediate host the Aphids were bred continuously 

 from May, several 3^ ^ appeared in due course, and in some cases oviparous 

 $ were also found. The (^ ^ are produced by alienicolae apterae toward the 

 end of summer on the intermediate host-plants, and the oviparous $ $ by 

 the physiologically specialised alienicolae alatae (sexuparae alatae) on the winter 

 host, at about the same time. The male-producing sexuparae are thus a.v. $ $. 

 One and the same mother may produce sexual males and sexuparae alatae, but 

 the sexuparae alatae only produce sexual females. 



It seems evident that in the adventures of migration the alienicolae alatae of any 

 generation may ahght on the winter host {Euonymiis) and produce young, resulting 

 eventually in colonies consisting of all stages, namely, a.v. $ $, w.v. $ $, 3^ cJ and 

 $ $. Mordwilko (1907) found all stages on Enonymus in Warsaw at end of September 

 1894. He also found a colony of a.v. $ 9 on Viburnmn opuhis in Warsaw at the end 

 of September. 



Similarly, sexuparae alatae alighting on intermediate host-plants may give rise to 

 sexual o $, which might thus be found together with colonies of agamic forms and 

 sexual males. This would explain the finding by Theobald (1912) of sexual ? $ 

 ovipositing on Rumex, by Gaumont (1913) on sugar-beet in October 1913, and by 

 Malaquin and Moitie (19i4) on haricot beans in October 1913. 



* It seems probable that during a mild winter agamic forms may persist throughout on certain 

 plants and carry on agamic reproduction normally in the following year. Davidson (1914) found 

 a colony of apterous agamic females on 30th January 1913 on Enonymus at Richmond. The 

 tree was taken into a greenhouse, and agamic reproduction was carried on normally throughout 

 1913. Several cases of long-continued parthenogenetic reproduction have been observed. The 

 question will be discussed in a later section dealing with the appearance of sexual forms. It may 

 be stated here that I have carried on a parthenogenetic strain throughout winter in a warm green- 

 house, winged sexuparae (which produce sexual 5$), sexual ^J^^ and parthenogenetic a.v. ^5 

 being produced in each generation from September to May. 



t Mordwilko (1907) was not able to obtain females on intermediate hosts. 



