598 MESSRS. O. THOMAS AND R. LTDEKKER ON THE [May 18, 



rirf5tly, it will not, we think, be contended by anyone, especially 

 in face oE the palaeontological evidence referred to below, that the 

 great number of the teeth of the Manatee has any direct connection 

 with the polyphyodontism of the primitive Mammalia recognized ^ 

 by many recent authors, even though Dr. Kiikenthal, in his account 

 of the embryonic distribution of the Manatee, says ^ : "Icli nehme 

 demnacli an, dass nicht weniger als drei auf einander folgende 

 Dentitionen sieh am Aufbau dieses Backzahnes beteiligen." 



Apart from " pre-lacteal " and " post-permanent " teeth, in whose 

 existence, with Messrs. Wilson and Hill, we should be glad to dis- 

 balieve, the utmost number that can be made out of the ordinary 

 mammalian set is 12, of which 4 would be milk-molars, 4 pre- 

 molars, and 4 molars. This is allowing for the possibility of the 

 milk-molars being regularly retained and the premolars coming 

 up behind instead of below them. Since, however, even with this 

 rather far-fetched explanation, the numbers are still far short of 

 the total required, we are disposed to think it unlikely, and prefer 

 to consider only the first three or four teeth as premolars, and the 

 rest as true molars. Whether such premolars belong to the per- 

 manent or to the milk series, we have no evidence on which to 

 base a suggestion. In Elephants, where the tooth-succession 

 is somewhat similar, the corresponding teeth belong to the milk 

 and not to the permanent series. 



If the presence of a specially large number of teeth in this 

 genus had any connection with a primitive multiplication of the 

 sets of teeth, the ancestors of Trichechus should have possessed an 

 equally redundant dentition, and on tliis point we are provided 

 with evidence to the contrary. For it fortunately happens that, 

 there are fossil Sirenians so closely allied to the modern ones that 

 we may almost treat them as if they were direct ancestors. 



Of these, by far the most important — because the best known — 

 is the Oligocene Halitherium, of which large numbers of speci- 

 mens have been described and figured by various authors, notably 

 Drs. Krauss ^ and Lepsius *. 



.In this genus a careful examination of the teeth seems to show 

 that although there was a distinct tendency towards the rapid wear 

 and degeneration of the anterior cheek-teeth so characteristic of 

 Trichechus, yet that the series of molars did not exceed four in 

 number, and in any case came to an end as soon as the animal was 

 adult. This latter point, so important for our present purpose, is 

 clearly demonstrated by Krauss's plate vi. and Lepsius's plate x. 

 fig. 96, where may be seen a terminal molar, considered to be m.*, 

 fully up, beginning to be worn, and yet without any trace of a 

 posterior tooth rising up to succeed it, as would be the case in the 

 Manatee. 



1 This, apparently with good reason, is altogether denied by the latest writers 

 on the subject, Messrs. Wilson and Hill, Quart. Journ. Micr. Sci. 1897, p. 427 

 et scqq. 



= Anat. Anz. xii. p. 524 (18%). 



3 N. Jahrb. Min. 1862, pp. 385-414, pis. \i. & Tii. 



* Abb. mittelrhein. geol. Vereins, 1882, pp. 100-200, pis. i. to x. 



