CHAP. IV.] 
ZOOLOGICAL LEGIONS. 
primary divisions of the earth. Neither can much stress be 
laid on the negative characteristics of New Zealand, since they 
are found to an almost equal extent in every oceanic island. 
Again, it is both inconvenient and misleading to pick out 
certain tracts from the midst of one region or sub-region and to 
place them in another, on account of certain isolated affinities 
which may often be accounted for by local peculiarities. Even 
if the resemblance of the fauna of Chili and Patagonia to that 
of the Palsearctic and Nearctic regions was much greater than it 
is, this mode of dealing with it would be objectionable ; but it 
is still more so, when we find that these countries have a 
strongly marked South American character, and that the north- 
ern affinities are altogether exceptional. The Roden tia, 'which 
comprise a large portion of the mammalia of these countries, 
are wholly South American in type, and the birds are almost all 
allied to forms characteristic of tropical America. 
For analogous reasons the Ethiopian must not be made to 
include any part of India or Ceylon ; for although the Fauna of 
Central India has some African affinities, these do not prepon- 
derate ; and it will not be difficult to show that to follow Mr. 
Andrew Murray in uniting bodily the Ethiopian and Indian 
regions of Mr. Sclater, is both unnatural and inconvenient. The 
resemblances between them are of the same character as those 
which would unite them both with the Paltearctic and Nearctic 
regions ; and although it may be admitted, that, as Professor 
Huxley maintains, this group forms one of the great primary 
divisions of the globe, it is far too extensive and too hetero- 
geneous to subserve the practical uses for which w T e require a 
division of the w'orld into zoological regions. 
Reasons for adopting the six RegioTis first proposed by Mr. Sclater. 
— So that we do not violate any clear affinities or produce any 
glaring irregularities, it is a positive, and by no means an un- 
important, advantage to have our named regions approximately 
equal in size, and wdth easily defined, and therefore easily re- 
membered, boundaries. All elaborate definitions of interpene- 
trating frontiers, as w r ell as regions extending over three-fourths 
of the land surface of the globe, and including places which are 
