84 MOSTLY MAMMALS 
smaller Argentine mammal, of the approximate size of a 
hare, named Pachyrucus. If it were not for the intermediate 
links, this creature would almost certainly be put down as 
a rodent, with which group it agrees in the structure of 
its teeth and toes, as well as in many other parts of the 
skeleton. Nevertheless, it is clearly a near ally of the 
typotherium, and therefore a member of the toxodon group. 
Here, then, we have one of the most remarkable instances 
of the phenomenon of parallelism in development. We 
have, in fact, displayed before us the origin of what we 
may call a rodent-ungulate: that is to say, an animal 
which, while certainly an ungulate by descent, has acquired 
such a marked resemblance to a rodent that, if we had not 
the intermediate links, it might be regarded as a member 
of the same order. This instance gives us some insight 
into the intricacies of evolution, and serves to show the 
amount of value attaching to many phylogenies of the 
animal kingdom. 
In addition to the slightly grooved huckle-bone, the 
toxodon group is characterised by at least one of the upper 
incisor-teeth growing throughout life, and by the cheek- 
teeth being either rootless or not forming roots till very 
late. There is, however a second group of allied extinct 
ungulates peculiar to the Argentine in which all the molars 
are rooted at the usual period, while the huckle-bone is as 
flat as in the elephants, although of somewhat different 
form. This group is represented solely by two genera, 
both of which are confined to the Patagonian deposits, 
where they are represented by animals rivalling rhinoceroses 
in size, and furnished with molar-teeth somewhat resembling 
those of the latter. One of these creatures, on which the 
name of Homalodontotherium has been conferred, presents 
the rare peculiarity of having the teeth arranged in a 
