98 Proceedings of the Royal Society of Edinburgh. [Sess. 
laboratory observations, the eldest-born of each generation being chosen as 
the parent of the next. The three garden series became extinct in August. 
In the nine weeks from June 2nd to August 10th they passed through 
five and four and five generations respectively, while during approximately 
the same time the A. I and B. I series in the laboratory passed through 
seven and five generations. I am inclined to account for their survival 
when the controls in the open disappeared, by the fact that they were 
better protected against enemies ; for the well-known disappearance of the 
red-currant aphis in the late summer seems to be partly natural, and due 
to the diminished birth-rate, and partly accidental, owing to the increase 
of insect enemies. 
The laboratory series passed altogether through fifteen of the eldest, 
and five of the youngest, generations. Eight of the A. I, 7 generation were 
transferred to Galeopsis tetrahit, and (including the cycle on the second 
host) this line of descent passed through eighteen generations. The B. I 
series passed through only eleven generations. Throughout the summer it 
was noticeably less productive than the other, and at the B. I, 5 generation 
there was a long period of inactivity. The parent— the apterous progeny 
of a winged female — delayed the fourth moult for over three weeks, and 
then died after producing three young at irregular intervals. 
Winged females may occur in the first generation, but my records 
show that the maximum of these forms appeared in the fourth generation 
in each eldest line of descent (fig. 8), but in A. IV and B. IV it occurred 
in the third and second respectively. From this it may be inferred that 
the appearance of winged individuals is not governed by position in the 
line of descent. On the observation bushes in the garden, the maximum 
appearance of the winged forms was some days earlier on green leaves 
than on red blisters. 
In mixed generations, the earlier births seem to be composed as a rule 
of both alate and apterous forms, while the later ones are nearly all alate. 
This rather supports the view that exhaustion of the parent may induce 
their appearance, but this is not borne out by the control lines (A. II, A. Ill, 
B. II, B. Ill), in which eldest and youngest were taken alternately in each 
generation. Mordwilko (22, p. 82) says that the appearance of winged 
females is accelerated by lack of food ; and this view is held by Davis 
(5, p. 132) for Aphis maidi-radicis, by Webster and Phillips (28, p. 82) 
for Toxoptera graminum, and by Buckton. As much evidence can be 
found against as for this theory, which rests on general statements and 
not on facts established by experiment. For instance, in M. ribis the 
maximum swarm of winged forms occurs just when food is most plentiful. 
