408 
ON THE OlllGlN OF 
try almost continuously connected with the peninsula of Ma- 
lacca, by means of Sumatra, Java, Timor, and other islands 
of the Indian Archipelago, 
The different races of our poultry are so remarkable for 
the extraordinary disagreement in respect to size, colour, 
and proportions, which they exhibit, that the reflecting na- 
turalist with difficulty convinces himself of their descent 
from one common stock. The smallest of the English 
Bantams do not weigh more than two pounds a-piece, 
whereas several varieties of eastern origin, such, for ex- 
ample, as the Great Cock of Java, sometimes exceed the 
weight of twelve pounds. In some the legs are long and 
bare of feathers, in others those parts are short and plumed 
to the very nails ; and, instead of the bare crown, or fleshy 
crest which distinguishes the greater number of kinds, the 
heads of certain varieties are surmounted by a large tuft of 
loose elongated feathers. I need scarcely add, that the va- 
rieties in the colouring of the plumage are infinite. 
These and other contrarieties, so greatly exceeding the 
utmost limits of whatever variation has yet been observed 
as the maximum of difference among the individuals of any 
clearly defined species, naturally lead us to inquire whether 
two or more species may not have been originally confound- 
ed under a mistaken identity of name and nature. Our 
knowledge of the natural history of the dog was long re- 
tarded by the erroneous views of Buffbn, and others, who 
argued that all the known varieties of that four-footed 
Cosmopolite, must be descended from the shepherd's dog; 
and although the usually acknowledged test of specific 
identity, that of two individuals of different sexes produ- 
cing a fertile brood, is supposed to have been conclusively 
applied in both instances, this I conceive does not so much 
prove that the multiform varieties which compose these 
:So called species, are respectively referable to a single stock? 
