16 ANATOMY 
knowledge of the anatomy of Birds, to decide which characters and 
which organs are of extrinsic taxonomic value, and which are 
not. Nor is it always possible to see why certain organs, fully 
developed, and exhibiting striking and constant features in one 
group of Birds, are extremely variable in another otherwise very 
circumscribed and apparently natural group. Supposing such a 
character to be absent in a given group, is it absent because it has 
not yet been developed, or is it because it has been lost ? Has it been 
lost by the ancestors of this group, or has it been abolished within 
this group? In the former case the absence of this character would 
probably help to -decide the relative position of the group; in the 
latter case this very same character would be reduced to a dia- 
gnostic point within the group, and not throw any light upon its 
relationship or systematic position. It may be very easy to dia- 
gnose genera or even large groups of birds, but this ability to deter- 
mine them by the help of mechanically arranged “keys” does not 
necessarily afford us more than an occasional glimpse of the sunk 
avine tree, at the reconstruction of which we all aim, as the true 
representation of the natural affinities of Birds. 
It is occasionally insisted upon that “tact” will help us to 
select and to reject characters, and thus prevent us from falling into 
glaring errors; but tact is a personal feeling, often bias, and it is 
proof, not inclination, that settles scientific questions. ‘The import- 
ance of these considerations, often expressed before in abler words, 
is gaining more and more ground among ornithologists, and will 
therefore permit the following illustrations of the ways in which 
we may or may not apply the study of comparative anatomy to 
classification. 
The presence of the AMBIENS Muscle is a Reptilian feature ; 
among Birds it exists in the majority of the lower groups, and is 
absent in most of the higher members of the Class. We conclude 
that the latter have lost this muscle, and not that it has not yet 
been developed in them. Its reduction or loss is still going on 
within some groups, such as Parrots and Pigeons. This loss takes 
place independently in widely different groups. It follows, first, 
that absence of this muscle does not always indicate relationship ; 
secondly, that we can derive forms that are without it from a 
group which still possess it ; but that the reversed conclusion is not 
possible. We know of no organ which has been redeveloped after 
it has once disappeared in the ancestors of the animals under con- 
sideration. Therefore the absence of the ambiens muscle in all 
Owls, which apparently use their hinder extremities in the same 
way as the Falconide (which possess this muscle), indicates that 
the Owls are not developed from the Falconidx, but from a group 
which, like the Macrochires, had already lost this organ. 
Similar arguments apply to the Caca. It is generally admitted 
