52 A Second Cooperative Study of " Vespa Vidgaris" 
range being from "954 to "996. There is, we think, no doubt that the absolute 
dimensions of the wings differ in the case of any single nest by as much as 1 to 4 p.c. 
from those of the general population. We have however still to bear in miud the 
possibility that this differentiation is not that of a nest against a general popu- 
lation, but may be that of a Surrey against a Buckinghamshire local race. 
Finally a third possibility is to be remembered, our spring queens may be severely 
selected as against the young autumn queens*. If this were so, the conclusion 
must be that the smaller queens have somewhat the better chance of surviving. 
The reader will observe that our results for left and right wings were closely in 
accordance, the maximum difference in the ratio of the means of the two series 
being in D, where it amounts to "006 ; whereas between different characters of 
the same wing it can be as lai'ge as ■041. This is sufficient evidence that the 
differences are not due to the same source. 
If we turn to the population queen, the right wing is larger for the characters 
A, C and G, and the left wing for the characters B, D, E and F. In the nest 
queen the right wing is larger for the characters A, D and G, and the left in B, 
C, E and F. The characters which have changed places are G and D. Actually 
E and F were the only characters in the nest queen and B and G in the population 
queen which showed significant deviations in right and left wings, owing to the 
size of the probable errors. But the general accordance between the two series 
seems to indicate that a real differentiation between right and left wing might be 
demonstrable with larger material. 
The last column of Table I shows that when we deal with the relative pro- 
portions of the wings, there is a far closer resemblance between population queen 
and nest queen than for absolute dimensions. The average ratio of the indices for 
the right wing for the two series differs from unity by only about 2 in 1000, and 
in the left wing by still less. Considering the probable erroi-s of the differences 
of the indices of the two series, most of these differences are undoubtedly significant, 
but, I think, we may assert that the degree of resemblance in pattern is greater 
than it is in size. 
The general conclusion to be drawn from examination of our 26 characters 
is that there is slight but quite definite differentiation between our single nest 
and our population of queens ; but that until further measurements are made we 
cannot on the basis of this table alone attribute the results without hesitation to: 
(a) local race differences, (6) individuality of the nest arising from heredity or 
nurture, or (c) the selective action of the winter iu differentiating autumn from 
spring queens "f. 
* Mr Latter suggests as a possibility that the " nest-queens " had fluid in their wings, while the 
" population queens " had theirs thoroughly dried by hibernation. It is not easy to determine how far 
this differentiation would be removed by pickling and again drying with the slide. But further obser- 
vations will be made in this direction. 
t We might even suppose that without a selective death-rate the wing might to some extent be worn 
in six months; the fact that the internal cells like/ and b are also smaller, appears to negative, 
however, such a hypothesis. 
