155 
blance and of difference no doubt may be traced through the entire 
structure ; as, for instance, the femur of the Proboscidia, although it 
wants the third trochanter, so characteristic of the more typical Peris- 
sodactyla, resembles the corresponding bone in that group in the cha- 
racters of the posterior side of its upper part. 
If it be admitted that this assemblage of singularly modified forms 
have sufficient resemblance to form a group which shall, with the 
more typical Perissodactyla, constitute two divisions, about equal in 
rank to the two divisions of the Artiodactyla, there cannot be much 
difference in opinion as to the manner in which this group should be 
subdivided into families. The Proboscidia stand forth as one (E/e- 
phantide), and the Sirenia as another (Manatide) ; while the Toxo- 
don, which in its most essential characters seems to agree with both, 
and in some points with each, has so many peculiarities of its own, 
that it appears entitled to rank as a distinct family of itself, which 
should be placed between the other two, not as a “ connecting link,” 
which its marked differences from either must forbid, and which if it 
were, it would but annihilate the distinction that exists. 
It seems time that naturalists should have decided what it is that 
constitutes an affinity ; whether a form can really be allied to several 
widely-different groups. We may naturally expect to find, that 
amidst the varied forms each part assumes, a character which is the 
rule among the members of one group may be the exception in an- 
other, without of necessity supposing that a species presenting such 
a character can truly belong to both, and thus tend to destroy the 
difference of the original models on which the two groups are or- 
ganized. In the present case, notwithstanding the peculiarities of 
structure mentioned as connecting the Toxodon with the Rodentia, 
its renowned describer, even while strengthening the idea of that affi- 
nity by adverting to Cuvier’s assertion that the Elephants approach 
the same order, yet places it, apparently without a doubt, among the 
Ungulata, to which it obviously belongs. Although Cuvier affirms, 
that if all the parts of the head of the Elephant be compared succes- 
sively with those of other animals, it is almost always among the Ro- 
dentia that their analogies will be found, he alludes only to three 
parts as indicating any such affinity. The relative size of the incisors 
and their alveoli can signify but little when their widely different 
structure is considered ; and he correctly tells us why the infraorbital 
foramen is large in both: the character of the os male is common 
to the Bats and Insectivora as well as the Rodentia, and seems to be 
a frequent concomitant of a degree of organization comparatively low. 
The direction of the incisors in the Toxodon differs very little from 
that which we find in many of the typical Perissodactyla, and the 
absence of roots is simply a physiological adaptation, and an indu- 
bitable proof that the detrition to which they were subjected was 
considerable ; while on the other hand, the whole structure of the 
cranium is on the ungulate type, especially different from the Capy- 
bara and the forms allied to it, whose skulls present so many striking 
characters, that if any resemblance really did exist, an anatomist to 
whom they were familiar would certainly perceive it at a glance. 
