Ernest Warren 
131 
the line AB to lie horizontally. When AB and CD were not horizontal slight 
adjustment of the cover-slip would almost always render them so and this was 
effected not through pressure exerted on the body but by alteration in the position 
of the body as a whole, for the legs of the animal projecting out in various 
Fig. 1. 
directions assured the necessary contact of the object both with the slide and the 
cover-slip. The antennae were as a rule very fairly straight ; when curved the 
bend was generally at one of the joints so that the measurement could be made by 
adding together two lengths (sec Fig. 1, G'E' -h E'D'). 
The absolute measurements AB {Frontal Breadth) and CD {Length of R. 
CD 
Antenna) and the ratio -^-g were all three dealt with in calculating the variability 
and the intensity of inheritance. 
(2) Growth. 
As the aphis grows the antennae become relatively longer, in other words the 
CD 
ratio .-^ rises. On this account it was essential, both for the ratio and the abso- 
AB 
lute dimensions, that all the parents and offspring should be measured as nearly 
as possible at the same stage of growth. In describing above how the experiment 
was conducted I used the words " mature female." Now, growth by no means 
ceases on the first production of young and I will define a "mature female "as 
one in which growth has become exceedingly slow. Growth at one period of life is 
very rapid, then it becomes slower and slower, but probably among many inverte- 
brates it never entirely ceases during the whole life of the animal. A mature 
female then is one which has arrived at that stage of life at which the period of 
