INTRODUCTION. IX 



general plan adopted by Prof. Flower ', although I have found it 

 necessary to introduce some modifications ; and in cases where good 

 working classifications of particular groups have been prepared by 

 specialists, such have frequently been followed. Thus in the Seleno- 

 dont Artiodactyla the classification of Prof. Riitimeyer 2 , of Basle, 

 has been adopted, while for the Perissodactyla Prof. Cope 3 is the 

 authority. A new scheme for the Artiodactyla and Perissodactyla 

 has recently been proposed by Dr. Schlosser 4 of Munich, to which 

 the reader's attention may be directed. This scheme adopts a phylo- 

 genetic system of families, and therefore differs very widely from the 

 systems generally employed, which run, so to speak, transversely to 

 such phylogenetic stems 5 . I have not seen the necessity for adopting 

 the order Bunotheria of Prof. Cope, since it appears to me that 

 their general primitive nature is the only bond of union between the 

 very diverse forms which are included in it ; and in this view I 

 have the support both of Prof. Flower and Dr. Schlosser. Neither 

 have I retained Professor Marsh's orders Pantotheria and Allotheria 

 for the Mesozoic Mammalia ; the majority of which appear so nearly 

 related to existing Marsupials that it has been a question whether 

 some of them should not be included in the modern families. In 

 his latest memoir Prof. Marsh 6 comes to the conclusion that the 

 Allotheria, which comprehend the Diprotodont genera, should be 

 regarded as at least a suborder of Marsupialia ; while the Panto- 

 theria, or Polyprotodont genera, are considered as being probably 

 placental 7 . 



1 In the 9th edition of the' Encyclopaedia Britannica,' in the Catalogue of the 

 Mammalia in the Museum of the Royal College of Surgeons, and in a paper 

 published in the ' Proceedings of the Zoological Society ' for 1883. 



2 Verb. nat. Ges. Basel, vol. vii. art. 2, pp. 29-60 (1883). In that memoir 

 the writer did not propose family names for the various groups, and it has 

 been necessary to accord as well as possible with such grouping. 



3 Proc. Amer. Phil. Soc. vol. xix. pp. 377-403 (1881). Amendments pro- 

 posed by the author of this memoir in later papers have been adopted. 



4 Morphol. Jahrb. vol. xii. pp. 1-136 (1886). 



5 See Cope, Amer. Nat. vol. xx. p. 720 (1886). 



6 Amer. Journ. ser. 3, vol. xxxiii. p. 345 (1887). 



7 The characters given by Prof. Marsh (op. tit.) as distinctive of the Panto- 

 theria do not differentiate them from the Marsupialia, and are not all applicable 

 to many of the forms. Thus in some examples of Perameles and Charopus the 

 canine has either a double or a distinctly grooved root (infra, p. 255) ; while 

 many of the English Mesozoic genera show a most distinct inflection of the 

 angle of the mandible. The dental succession of Triconodon (infra, p. 257) 

 appears both to Mr. O. Thomas and myself to be absolutely conclusive evidence 

 of its marsupial affinities. 



