130 
Variation and Inheritance in Aphis 
The weather at that time was very warm and dry but the air around the 
plants was kept moist by constantly drenching the table and the surrounding 
ground. On watching from day to day the widely scattered and perfectly quiescent 
individuals one could see the young being born, and when a little cluster of 6 — 12 
young had been produced by a mature female the mother was carefully taken off 
and preserved in absolute alcohol. The brood on the leaf was then enclosed in a 
little cage of bolting-cloth which was firmly tied around the petiole. On previous 
occasions I have used cylinders of very thin glass with muslin tied over the top 
and bottom : in these cages, however, there was so much condensation of water 
that the animals were often drowned. Bolting-cloth cages answered exceedingly 
well, they stand away from the leaf, the observer can see through them, and there 
is no injurious condensation. 
After about a dozen families had been caged in the manner described the 
plants had all become so infested that it was not safe to assume that any given 
cluster of young had all been produced by the adjacent mother. After this, a 
mature female was placed on a carefully cleaned leaf and enclosed in a cage. 
When the individual had produced a batch of young it was removed and preserved 
in alcohol and the brood was allowed to grow up. I believe the cultures were 
absolutely pure, for the nvimber of young enclosed in the cages was always counted 
and registered, and afterwards compared with the number in the adult broods. 
When the brood had become mature, grandchildren of the original parent had 
been produced, and the adult individuals were preserved. Often by this time the 
leaf had become somewhat faded, and when such was the case the leaf with the 
grandchildren was cut off and placed on a fresh leaf and the whole enclosed in a 
cage. Frequently under such circumstances the whole brood died, for the young 
did not always seem capable of migrating to the adjacent leaf Out of 60 broods 
I only obtained 30 sets of grandchildren. 
The Measurevients. 
The dimensions selected for measurement were: (1) the distance between the 
eyes AB (Fig. 1) and (2) the length of the right antenna CD. Body-lengths such 
as XY or AZ were unreliable owing to the softness and consequent contractility of 
the hinder region of the body ; sometimes the segments comprised in ZY were 
more or less telescoped inwards. 
The specimen in a drop of spirit was placed in a well-slide and covered 
with a small piece of thin cover-slip and the dimensions were read off by an 
ocular micrometer under an A Zeiss objective. The magnification was such 
that the measurements were made to the accuracy of '00221 of a millimetre*. 
Fortunately the shape of the animal is such that there was a strong tendency for 
* Subsequently for the purpose of calculating the constants it was found that the variability was so 
considerable that the unit employed was •0221 or -0442 mm. 
