662 oe W. C. Crawley on 
worthy that not a single ¢ has been produced in either 
colony. 
The further history of this colony, together with that of 
the former one, tends to show that the egg-laying capacity 
of the Gs is not of very long duration, lasting two seasons 
at most, while the life of the § may be from three to seven 
years. 
The first eggs of this year were laid on May 10th, six- 
teen days earlier than in 1910, and the first larva (from 
last year’s eggs) reached the pupal stage on May 30th, and 
on June 7th there were about fifty pupae. On July 11th 
I found four callows in the nest, all of a yellow colour, and 
unmistakably Z. wmbratus. On the 15th there were over 
thirty. By July 27th this number had increased to over 
one hundred, and there was no sign of any hostility towards 
them on the part of the niger Os. 
It is curious that while the young wmbratus $s devoted 
their attention almost entirely to the care of the larvae, the 
niger Ss alone surrounded the wmbratus queen. 
At the present moment (October 12th, 1911), the 
umbratus Os, as far as can be judged, shghtly outnumber 
the nigers, and the latter still pay more attention to the 
queen than the ants of her own species. Not a single 
niger has hatched in the nest this year. From the number 
of the larvae, presuming these tu be exclusively wmbratus, 
as is almost certain, the niger population of the colony will 
be outnumbered by at least four to one next year. 
Allowing for the increased mortality among the more 
active niger in nature, it would seem that it takes four 
to five years for a colony founded in this way to become 
exclusively wmbratus. In this particular colony the total 
number of deaths from natural causes among the niger 
from September 1908 to September 1911 was 416, and 
there are still over 800 niger Ss alive in the nest. 
It has been suggested by Wheeler (1909) that the usual 
sterility of ds is influenced by their being constantly in a 
state of semi-starvation, nearly all the food they obtain 
being given to the larvae. Thus in the second colony 
(No. 2) the workers were without larvae to tend from July 
to October 1908, and therefore all the food was consumed 
by them. But in the first colony (No. 1) this was not the 
case, as the nest contained larvae from the first. 
It appears also that in certain species when a colony is 
deprived of its queen several workers become fertile. It 
