88 
Proceedings of the Royal Society of Edinburgh. [Sess. 
species. Here the descendants of two stem mothers were followed down in 
each case through four lines of descent. Six of these lines of both stocks 
were fed entirely on green leaves, but the seventh and eighth (by a 
coincidence in both instances it was the youngest of both series) were fed 
on blistered leaves. The result was that all the latter had the red blister 
type of antenna, while all the rest, irrespective of generation, had the green 
leaf type. That position in the brood had nothing to do with the modifica- 
tion is shown by the A II, B II, A III, and B III lines of descent. Here 
Fig. 5. — Curve showing the relation of the 
permanent sensoria to the articulation 
of Joints V and VI in nine females of red 
blister stock reared for three generations 
on green leaves. 
The abscissae represent the difference in — h_ mm. between the sensorium and the articula- 
100 
tion. The lower curves refer to the fifth, and the upper to the sixth, joint. 
(see generation series) an attempt was made to check the eldest of eldest 
and youngest of youngest lines by choosing alternately the eldest and 
youngest of the brood as parent to the next generation. 
All the winged forms of these four “ control ” lines of descent have the 
green leaf type of antenna. 
The question arises what is the cause of the modification of the antenna 
in this species ? Taking as a working hypothesis that it is due to some- 
thing in the food, I tried the effect of rearing thirty-eight young of a 
female of red blister stock on green leaves from their first instar ; but the 
result was on the whole negative, although the range of variation in the 
