Parthenogenesis in Worker Ants. 659 
were laid by the queen in 1896 (in my experience the 9s 
of L. umbratus do not lay till the year following impregna- 
tion), and the niger larvae left by the old queen lived 
through the winter and began to change into pupae at the 
beginning of June 1897. On June 26th the wmbratus 
queen began to lay, and by August Ist there was a large 
quantity of eggs laid by her” (and also, as the results show, 
by the niger 9s). “The larvae from these eggs lived 
through the winter. On May 31st, 1898, the first larva in 
the nest changed into a pupa, and by the end of July 
there were about one hundred pupae in the nest. On 
August 2nd twenty pupae hatched, the young ants all 
being L. niger Os.” This result being inexplicable to me, 
as I took it for an established fact that if eggs had been 
laid by the niger 9s the resulting perfect insects would 
have been fs, I tried to explain it by assuming an error in 
my records, as follows :—“ It is clear that there is an error 
in my records here, for it is hardly possible that larvae 
from niger eggs laid in 1896 could have lived till 1898 
before completing their metamorphoses. I have recorded 
giving the nest some niger pupae on August 6th, but some 
must have been given to the nest before August 2nd.” 
However, in view of the identical results obtained with the 
other similar colony from 1908 to 1910, where there was 
no possibility of error, I think that there was no mistake 
here, and that the results were due to parthenogenetic eggs 
laid by the niger 9s producing 9s. The following year, 1899, 
hundreds of L. wmbratus $s came to maturity in the nest, 
and were all, as stated above, killed by the niger 3s. No 
niger OS appeared after 1898. In 1900 the wmbratus 9s 
began to hatch in July, and this time were not molested by 
the nigers, and by July 18th there were twenty alive and 
carrying pupae about the nest. This interesting colony 
unfortunately came to an untimely end, as I was obliged to 
leave it in other hands for more than two months, and 
found on my return all the ants dead except the queen, 
who died shortly afterwards. There was not a single niger, 
or remains of one, to be found in the nest, though there 
were hundreds of dead wmbratus 8s. 
During the four years this colony was in my possession 
not a single f of either species made its appearance. 
Again, in September 1908, I had a queenless colony of 
L. niger which accepted a fertile wmbratus 2 as queen on 
September 19th. I was determined that there should be 
