meen YAS, THE USURPER-BEE 69 
the cocoons of all the species that I have observed 
(vupestris, vestalts, distinctus, and barbutellus) differ 
from those of their Somdus hosts in that they 
quickly lose their fresh lemon tint, and become of a 
dull ochreous colour, and later turn semi-transparent, 
crackling when they are dented—qualities that are 
possessed in only a very slight degree by the 
Bombus cocoons. The cocoons of Ps. rupestris 
form compact clusters like those of its host B. 
lapidarius, while those of Ps. vestalis, like those 
of L. terrestris, are only loosely attached to one 
another. 
The Pszthyrus kills the Loméus queen before 
she has laid the full number of worker eggs, con- 
sequently nests containing Pszthyrt are not very 
populous, the number of workers seldom exceeding 
eighty. In nests of B. dapidarius containing Ps. 
rupestrtis | have never known a /apizdarzus queen 
to be reared; similarly in nests of B. terrestris 
attacked by Ps. vestalis | have never seen a young 
terrestris queen. 
In a Pszthyrus-ridden nest a large number of the 
workers become fertile. In digging out a nest of 
B. terrestris, 1 can generally tell before reaching 
the nest that it has been victimised by Ps. vestalis if 
many workers containing eggs, known by their 
shining black abdomens with the yellow band on 
the second segment more or less obliterated, rush 
out of the hole to greet me. But it is a remarkable 
fact that as long as the Pszthyrus queen reigns in 
the nest I have not known a Lomdbus male to be 
