330 
CRESSWELL SHEARER. 
In D. gyrociliatus parthenogenesis is only brought 
about by experimental means, tliat is by cutting out the male 
eggs from the capsules, inside of which they have been 
deposited in company with the female eggs. 
As early fertilisation of the female before she leaves the 
egg-capsule is invariably the rule in this species of Dino- 
philus, parthenogenetic development can only take place in 
the complete absence of the male. This I believe never takes 
place naturally, for during the course of three years* work 
I have not met with an egg-capsule that did not at least 
contain one male in addition to the female eggs. The propor- 
tion of three female to one male egg, or somewhat less than 
three to one, is remarkably constant, as shown by the table in 
the following section, which is made up from capsules collected 
at various seasons of the year. Parthenogenetic develop- 
ment, therefore, can only be brought about by the actual 
cutting out of the male eggs. Eggs in which early fertilisa- 
tion has been prevented in this manner, after a certain period 
develop parthenogenetically, and the females to which they 
give rise in turn lay eggs of two sizes, male and female. 
The sex-determination taking place in the last, therefore, is 
that which I have called the sex-determination of the 
parthenogenetic cycle. 
I have not, however, yet satisfied myself completely that I 
have succeeded in entirely eliminating the males in all these 
cases. This has to be done in the later segmentation stages 
to make sure of the matter. The extreme stickiness of the 
living egg, coupled with the small size of the male egg, 
render this a very difficult operation to perform. It is 
nearly always attended with failure. The small male eggs 
readily escape observation or adhere closely to the large 
female eggs and so make fertilisation possible at a later stage. 
For these reasons I feel considerable doubt if parthenogenetic 
development ever takes place, and I wish to reserve my final 
opinion on this subject till I have tested it by a larger 
number of experiments. 
In one or two cases, however, where I have taken excep- 
