134 THE RUDDY SHELLDRAKE OR BRAHMINY DUCK. 
specimens, a sort of barred appearance is this produced on 
all the parts above referred to, while in some, where the pale 
tippings are very broad, the colour of the bird seems altogether 
washed out. The result is a very great difference in the 
appearance of specimens. This pale tipping cannot be an invari- 
able seasonal change, as we have many pale or pale mottled 
birds killed at the same time and place as fully coloured ones, 
nor can it, I think, be solely due to differences in age, since 
during December and January, the birds showing more or less 
of the pale tippings greatly predominate, while they are com- 
paratively rare in March. The point is one that requires 
investigation. 
Very commonly the white wing-coverts are entirely over- 
spread in the male with a paler or deeper, richer or duller, 
orange buff shade. Indeed I have no male killed before the Ist 
of April in which ¢races of this are not visible, and in one 
or two birds it is so strongly marked that the wing-coverts are 
unicolorous with the breast (in all these examples, itself very 
pale.) But in the female, killed during the same months, the 
wing-coverts are more generally nearly pure white; still a good 
many even of the females show traces of the orange buff 
shade on the coverts, and I have one specimen in which they 
are nearly as richly coloured as in any male. 
In some birds the speculum is a deep green, in almost all 
lights ; in others it is aimost always a deep purplish bronze. 
In some birds the paler colour of the head and upper neck 
is abruptly defined all round against the richer tint of the 
lower neck, while in some the one colour passes by insensible 
degrees into the other. 
I may add that in the majority of birds the feathers of the 
lower abdomen are deeper coloured and more of a chestnut than 
the rest of the lower plumage, forming a large and often very 
conspicuous patch; but in some, probably the birds of the 
year, there is no trace of this. 
A nestling brought from the Tso-mourari is mostly white, 
marked on the upper surface with blackish brown, and with 
here and there a fulvous tinge. 
To THE RESTRICTED genus or sub-genus Casarca (as indicated, 
note p. 123,) belong three other fine species—C. cana, long con- 
founded with our bird, from South Africa, C. tadornoides from 
Tasmania, Southern and South Western Australia, and C. varie- 
gata from New Zealand, perhaps the handsomest of all. 
Cre 
