364 
The Egg of Cumins Canorus 
worth while to test the point in question in respect of the dimensions of length and 
breadth. The method employed is fully explained in my previous paper, but it will 
be convenient to re-state it. The mean {M) length or breadth, as the case may be, 
of all the Cuckoo's eggs is first computed, thence the standard deviation (cr) is 
obtained by the formula 0-2 = ^^^^-lil^ — ^ ^ where x = the measurement of any one 
egg and n = the number of eggs measured : the coefficient of variation, y 
then be found. To test whether the deviation occurring in Cuckoo's eggs deposited 
in the nests of any particular species of foster- parent is significant, M,. is taken as 
the mean of the whole race of Cuckoos, and Mg as the mean of Cuckoo's eggs 
found in the nests of the foster-parent species : the standard deviation {a^) of such 
eggs is also ascertained. The value of — Mg is then compared with that of 
0'67449 A / -f- " (where — total number of Cuckoo's eggs, and n.^ = the number 
of Cuckoo's eggs in the nests of the species in question), which is the probable 
error of M,. — Ms due to random sampling. If the value of if,. — Mg be not at least 
2 to 3 times as great as the value of the other expression, then the difference of 
Mr and Mg is not definitely significant. It will be seen by reference to the 
tabulated summary below that in the case of four of the sixteen sets examined, viz. 
Robin-Cuckoos, Wren-Cuckoos, Whitethroat-Cuckoos, and Hedge-Sparrow-Cuckoos, 
the differences are significant in respect of both length and breadth. Of the 
remainder, five, viz. Garden-Warbler-Cuckoos, Barred- Warbler-Cuckoos, Marsh- 
Warbler-Cuckoos, Reed- Warbler-Cuckoos, and Tree-Pipit-Cuckoos, exhibit differ- 
ences significant in one dimension, but doubtful in the other; two, viz. Sedge-Warbler- 
Cuckoos and White-Wagtail-Cuckoos,exhibit significant differences in one dimension 
only; one, viz. Meadow-Pipit-Cuckoos, exhibit doubtfully significant differences in but 
one dimension ; and four, viz. Redstart-Cuckoos, Red-backed-Shrike-Cuckoos, Pied- 
Wagtail-Cuckoos, and Yellow-Ammer-Cuckoos, do not present differences signifi- 
cant in either dimension. The balance of evidence is thus decidedly in favour of 
there being distinct sets of Cuckoos, As a result it now seems fairly certain that 
Prof A. Newton's suggestion {Dictionary of Birds, p. 123) is correct, and that there 
are certain " gentes " of Cuckoos whose members being closely related lay eggs of 
somewhat similar dimensions, and in the main confine their attentions, generation 
after generation, each to its own particular variety of foster-parent. In other 
words, the evidence here adduced points to the conclusion that the species Cuculus 
canorus is tending to break up into a number of sub-species, each with its own 
particular habits in respect of selecting a foster-parent. There are recorded 
numerous instances of individual Cuckoos adhering to one species of foster-parent 
whenever possible, but it appears that in the event of failure to find a nest of the 
species preferred the Cuckoo places her egg in any nest that comes handy. This 
fact probably accounts for the very large number of species in whose nests Cuckoo's 
eggs have been found, and emphasises the importance of obtaining large numbers 
of measurements for investigations of this character. 
