Germ Cells and early Embryology of certain Aphids. 615 
to the willow twigs by means of this sticky elastic capsule. It is 
very difficult to remove the eggs from the willow twigs immediately 
after deposition without rupturing them, but after being exposed 
for twenty-four hours to the air, or after fixation, the eggs can be 
removed very readily from the willow twigs. 
The winter eggs begin to develop immediately after deposition. 
The embryo passes through the winter in a half-grown condition 
and is situated within the center of the egg, being completely sur- 
rounded by the yolk. Growth is completed in the following spring. 
There are, however, exceptions to this rule. Embryos need not reach 
half their normal size in the fall in order that growth may be 
continued in the following spring. If the embryo has reached the 
germ band stage, i. e. when the germ bond is completely separated 
from ‘the serosa (Pl. 51, Fig. 51) and lies completely within the 
center of the egg, growth may be completed in the following spring. 
But, on the other hand, when the blastoderm is completely formed 
and only partly invaginated, the embryo does not survive the winter. 
A very small per cent. of the late deposited eggs develop in the 
following season. Eggs deposited on willow twigs in the green- 
house begin to develop like those in favorable conditions out of 
doors, but do not hatch out. Those brought into the green-house 
immediately after one or several freezings will hatch out. Thus it 
appears that freezing is necessary for the completion of development. 
In studying the life history of the aphids we find that partheno- 
genesis and the determination of sex are directly related. It is 
quite evident that fertilization does not play a direct role in the 
determination of sex as found in some of the Hemipterean insects where 
the chromosomes vary in size and number in the different germ 
cells. If it did we would expect to find two distinct parthenogenetic 
lines beginning at the time of fertilization of the winter eggs. In 
the one the paternal characters would dominate throughout the 
parthenogenetic series and give rise to males only in the fall of the 
year when the sexual generation appears. In the other, the female 
characters would dominate and give rise to sexual females only. 
We were unable to find a single instance where a given stem mother 
gave rise directly to either a male (paternal) or female (maternal) 
line. The first evidence we do have of sex being separated, i. e., 
a dominance of either male or female characters, occurs in the 
fifth parthenogenetic generation where a single parthenogenetic 
individual produces presexual forms that give rise to either all 
