322 ANALYSIS OF THE GRALLZE. 
these may be regarded as belonging to the terrestrial 
and fluviatile portion of this class there is a pre- 
ponderance in favour of the marine portion. With 
respect to very many of this class, it is quite im- 
possible as before remarked, to assign to them un- 
equivocally an inland or marine station. ‘They are 
either so imperfectly known that their preference is 
merely suspected by their accidental occurrence in 
one of these situations on the occasions when seen, 
or they have been found to resort equally to both 
kinds of habitats ; even in the case of those species 
which breed on our moors, and appear in autumn 
and winter on the coasts, allowing that they do not 
at those seasons visit lakes and rivers also, it must 
cause a doubt on the mind whether it would be 
natural to refer them to the inland or to the marine 
class exclusively. In fixing the numbers however 
as above, at twenty-two for the cultivated districts, 
and the remainder thirty-four for the uncultivated 
parts and the shores I have been guided thus: 
I first selected those respecting whose station no 
doubt could be felt, and then classified the rest by 
ascertaining as far as possible which parts of South 
Devon they evinced the most preference for, or in 
the case of the rarer birds by considering that to be 
their station where each had by good accident been 
observed, unless it seemed to me that this station 
was at variance with the general character of the 
bird and assumed only by mere casualty. Of the 
thirty-four then, I have reckoned two as belonging 
to Dartmoor, and thirty-two as shore birds. ‘Then, 
besides the two which are peculiar to the moor, 
namely the Great Bustard and Crane, more than 
two dozen other species have been noticed on that 
spot, very many of which breed there. Some of 
these I have considered as belonging to the culti- 
vated parts and some to the shores, according to 
the bias they betrayed. Again, of the twenty-two 
