SYSTEM OF NATURE. 
43 
of generation is in many respects strikingly similar. The 
differences appear to be — 1st. Those of external appear¬ 
ance. 2ndly. That the anterior limbs of birds are fitted 
for flight, instead of terrestrial progression : even this im¬ 
portant difference is not abruptly accomplished : if we 
carefully examine the structure of the kangaroo, we shall 
see that its fore-feet are diverted from their normal employ, 
and assume the appearance and office of hands rather than 
of feet; its hind-legs also are altered in their character, 
and have assumed a considerable similarity to the legs of 
the ostrich and other walking birds. 3rdly. That the fea¬ 
thers of birds are substituted for the hairs of marsupials ; 
and here again the marsupial form is, as it were, a connect¬ 
ing link, for even their hair is not that of the placentals, 
but seems of intermediate character between hair and fea¬ 
thers. 4thly. The nest or site of incubation differs: in 
birds it is external and artificial; not attached to, and a 
portion of, the body, as in marsupials. Hence it appears 
that marsupials and birds, considered as a double group, 
may be distinguished from all other vertebrates by the 
single fact of their incubation; from all oviparous verte¬ 
brates by their warm blood ; and from each other by the 
respective characters of pouch-bearing and nest-making. 
Nothing could be more easy than to reduce the class of 
birds into seven minor groups. Both llliger and Latreille 
have made a septenary arrangement of them, and each of 
these arrangements has usually been considered good and 
natural. An objection, however, occurs to me, which I think 
demonstrates the necessity for hesitating a moment before 
adopting any numerical division, unless obtained by some 
uniform and logical process. The objection is this : twenty - 
one authors of reputation have proposed as many systems 
