74 
ternal appearance that may be noticed among its members, so that 
we should anticipate but little difficulty in subdividing the order 
into a number of natural groups, the confusion, and differences of 
opinion that have existed, not only as to the manner in which the 
order should be divided, but also as to the position which certain 
forms should occupy, show sufficiently that the task is by no means 
an easy one; and when the structure of the different members of 
the order is investigated, and those forms are known to us by which 
the most strikingly different genera are blended one into another, it 
becomes difficult to draw the lines of separation, and still more to 
fix the characters by which the groups can with accuracy be distin- 
guished from each other. In the present state of zoological science, 
it seems scarcely worth while to allude to the distinction of planti- 
grade and digitigrade, which though due to no less an authority than 
Cuvier, can hardly be said to possess any claims to the title of a 
philosophic distinction. Indeed the former of these divisions, if the 
character be fully insisted on, will include a very incongruous assem- 
blage of forms. 
It is upon the differences of the teeth that the subdivision of this 
order has been made chiefly to depend; but, although it does so 
happen that in most cases the affinities of a species may be truly 
predicated by the inspection of these organs, there are some in which 
naturalists have been led into error by too rigidly depending on 
them; it must be recollected that, especially in an order like this, 
where we find among the different species, every gradation between 
a purely carnivorous diet, and the capability of subsisting entirely 
on vegetables, the teeth, by the various degrees to which the differ- 
ent cusps are developed, and also by the point at which the normal 
development of true molars from behind may be arrested, present a 
very great variety in the amount of tubercular surfaces, or of tren- 
chant edges, to suit the regimen of the species, without any neces- 
sary connection with its true affinities. For instance, the remark- 
able variation in the number of true molars presented by the different 
genera of the Dog-tribe is known to naturalists; and my own col- 
lection possesses the skull of a small dog in which, such is the 
arrest of development resulting from the shortening of the jaws, 
that although the individual was very old, it had never developed 
mere than one true molar above and two below, or one behind the 
carnassial tooth in each jaw, being one less than is usual in the 
species. 
If we except the aberrant family of Seals, we find that this order 
does not present so many of those very striking extremes of adaptive 
modification as are to be met with in some others, the generally 
lithe and active form prevailing through the order rendering a very 
moderate amount of adaptive modification necessary to fit the animal 
for almost any situation and mode of life, and from this cause it 
also happens that since the fallacious nature of the old division into 
plantigrade and digitigrade has been generally perceived, the classi- 
fications of this order most usually adopted by naturalists have ap- 
proached much nearer to those natural divisions, which the essential 
