xii THE ORDER BRUT A 161 



by every beginner in Zoology, and all the facts which have 

 already been accumulated relating to the numerous extinct 

 forms of Proboscideans, whether Mammoths, Mastodons, or 

 Dinotheria, were quite unknown to Linnseus. One species 

 only, Mephas maximus, represented in the zoology of a hundred 

 years ago all that was known of the elephants or elephant-like 

 animals. 



The genus TricJiecTius of this edition exhibits a very curious 

 phase of zoological knowledge. It contains two species 1. 

 T. rosmarus, the walrus, now known to be a modified seal, 

 and therefore a member of the Linnsean order FER^E, and 2. 

 T. manatus, a name under which were included all the known 

 forms of manatees and dugongs, in fact the whole of the 

 modern order Sirenia, animals widely removed in all essential 

 points of their organisation from the walrus, with which they 

 are here generically united. Their position, however, between 

 the elephant on the one hand and the sloths on the other, 

 is far better than their association with the Cetacea, as in 

 Cuvier's system, an association from which it has been most 

 difficult to disengage them, notwithstanding their total dis- 

 similarity except in a few external characters. Although the 

 discovery of many fossil forms has done much to link together 

 the few existing species and to show the essential unity of the 

 group, it has thrown no light upon their origin, or their 

 affinities to other mammals. They still stand, both by their 

 structure and their habits, in a strangely isolated position, and 

 it baffles conjecture to say whence they have been derived, or 

 how they have attained their present singular organisation. 



The remaining genera of the Linnsean order BRUTA con- 

 stitute the group out of which Cuvier, following Blumenbach, 

 formed his order Edentata, a name certainly not happily 

 chosen for a division which includes species like the great 

 armadillo, having a larger number of teeth than any other 

 land mammal, but which, nevertheless, has been so generally 

 adopted, and is so well understood, that to attempt to change 

 it would only introduce an element of confusion. Four out 

 of five of the principal modifications of form in the group at 

 present known are indicated by the four Linnsean genera, 

 Bradypus or sloth, Myrmecophaga or ant-eater, Manis or 



M 



