Mr. G. R. Waterhouse on the Classification of Mammalia. 403 



Manatus, &c.) differs as widely from that of the carnivorous 

 Cetacea as do their habits : that the amount of variation is as 

 great as well could be in animals of the same class, existing in 

 the same great deep. The junction of the Dugongs and J\la- 

 natees with the true AATiales cannot therefore be admitted in a 

 distribution of animals according to the organization. With 

 much supei-ficial resemblance, they have little real or organic 

 affinity to the Wah'us, which exhibits an extreme modification of 

 the amphibious carnivorous tjqoe. I conclude, therefore, that the 

 Dugong and its congeners must either form a group apart, or be 

 joined, as in the classification of De Blainville, lAdth the Pachy- 

 derms, with which the herbivorous Cetacea have the nearest 

 affinities, and to which they seem to have been more immediately 

 linked by the now lost Dinotherium." 



On the whole then it appears to me, that the researches of the 

 author just quoted, and of De Blainxdlle, will bear out the as- 

 sumption, that the animals forming the family of herbivorous 

 Cetacea in the ' Regne AnimaP are in fact aquatic Pachrjdermata, 

 bearing the same relations to the ordinary Pachyderms as do 

 the Seals (of which group the Wah'us forms part) to the Carni- 

 vora. Tliat there is a strong analog}^ between these animals and 

 the true Cetaceans cannot be denied, but that there is any direct 

 affinity I think is doubtful. 



In the circles representing the different orders, I have intro- 

 duced those genera belonging to each which appear to approach 

 most nearly to other orders. Most of these approaches of genera 

 of one order to the general characters of other orders have been 

 before pointed out ; I cannot pass on, however, without making 

 some observations upon the natui'e of these approaches. Exam- 

 ples of this kind are numerous, and have given rise to a common 

 beUef, that, as a general rule at least, the various sections of 

 animals, even those of the highest value, are gradually hnked 

 together. It has been most frequently stated, that the groups, 

 large and small, of which the animal kingdom is composed, 

 blend imperceptibly into each other ; and supposing this view to 

 be correct, it follows that there are many species so well balanced 

 in their characters, that they cannot in a classification, without 

 doing \iolence to those characters, be placed in any particular 

 order ; these Hnks must be aiTanged between the orders, the 

 characters of which they combine. But in those groups to which 

 I have paid most attention, I will venture to assert, that species 

 which even appear to require to be so located, are far from being 

 numerous, and moreover, that in proportion as knowledge of the 

 groups and species increases, so does the nvunber of supposed 

 links decrease ; that is to say, it becomes less and less doubtful 

 as to the group in which an animal should be placed. A short 



