April, 1904.] 



KNOWLEDGE & SCIENTIFIC NEWS. 



63 



agree in their divided scaphoid and lunar. As indi- 

 cated by the name of the typical ijcnus, the ]'ii(ii(iviJu- 

 are re,t;arJed by .\nierican pala-ontoloi^ists as the ances- 

 tors of the ci\ets (r/rvM/iL-) of the old world; and it is 

 not improbable that they were likewise ancestral to the 

 primitive dogs. If it be added that there is evidence to 

 show that the members of the weasel tribe are also 

 sprung from a more or less nearly allied Eocene group, 

 we shall have accounted for the origin of four of the 

 most important families of existing land Carnivora, 

 namely dogs, bears, civets, and weasels. As regards 

 hya?nas, there is little doubt that they are closely related 

 to civets, with which they appear to be connected by a 

 number of extinct forms, such as IitHI:eiiiiw. 



Leaving the raccoon family alone, it may be added 

 that there is still some degree of uncertainty with regard 

 to the origin of the cats (Felida-). Unless, l)owe\ er, they 

 trace their origin direct to the undermentioned creoclonts, 

 there seems to be considerable probability that they 

 are derived from the imperfectly known family of primi- 

 tive camassidents termed PaLconklidte, all the members 

 of which are characterised by their short jaws and cat-like 

 dentition. In the typical Palteonictis, which dates from the 

 \\"asatch, or Lower, Eocene, the carnassials are somewhat 

 imperfectly differentiated from the other teeth ; but in 

 .Ulurotherium of the Bridger they become well characterised 



Having thus traced, more or less definitely, most of 

 the principal families of existing land Carnivora to 

 generalised forms which are evidently on the bordeiland 

 between the Carnassidentia and some more primitive 

 type of Carnivora, we have to turn our attention to what 

 is known with regard to the latter. 



WU*V^ 



Fig. 4.— 5kull of Sinop.:, a North American Creodont. (After Wortman.) 



Such primitive type is represented by the Eocene and 

 Oligocene Carnivora collectively known as Crecdontia, 

 of which the American'5(H(?/'a or Stypolophus" (fig. 4) and 

 the European Hycenodon and Pterodon (fig. 5) are well- 

 known representatives. In addition to other features 

 which cannot be noted here, these creodonts are collec- 

 tively characterised by three long, narrow, small-brained 

 skulls, by the fact that the scaphoid and lunar of the 

 wrist are usually distinct, and, above all, by the non- 

 development of a pair of differentiated carnassial teeth. 

 In place of these, the lower jaw (fig. 5) has all the three 

 molars of a cutting type ()«', )«-', w;^) ; and it will be 

 further noticed that these teeth differ from the corre- 

 sponding teeth of a carnassident by the circumstance 

 that they increase in size from the first to the third, 

 instead of decreasing. These animals all have five- toed 

 feet, in which the thumb and the first toe may he 

 opposable to the other digits. 



Unless these creodonts have given rise to the true seals 

 of the present day, they seem all to have died out during 

 the Tertiary period without leaving any descendants. 

 Moreover, they appear to have been derived from some 

 still more primitive stock independently of the carnas- 



* Represented in Europe by ihe closely allied Cynohyctnodon. 



sidents, with the earlier forms of which latter they were, 

 however, evidently allied. In other words, camassidents 

 and creodonts appear to be (h\eri;ing branches from a 

 single primitive stock-, which proh.ibiy li\t\l iluring the 

 Secondary, or Mesozoic, epoch. 



^'^k: AJ;:-d'%f 





FiK. ;;.• 



-Lower Jaws of Creodonts and Marsupials. 



2 PtfTudon, 3 H(trU;i'i lift, 4 TiiijlacintiH. 



What were these Mesozoic ancestors, is the next ques- 

 tion which presents itself. 



.V comparison of the lower javvs of the creodont 

 Hyctnodon and Pterodon with that of the marsupial 1 hyla- 

 cinus, as displayed in fig. 5, shows at a glance that the 

 dentition in all three is of the same generalised type ; 

 this being especially indicated by the form and relative 

 dimensions of the three molars. It is true, indeed, that 

 in the marsupial there appear to be four of these teeth ; 

 but this is due to the fact that the tooth in advance of 

 these (m.p. 4) is a persistent milk-tooth, which is not 

 replaced, as in the creodonts, by a permanent premolar 

 (pp. 4). Certain South American extinct types such as 

 Borhyana (fig. 5 — 3) are intermediate in regard to the 

 number of teeth replaced between creodonts and mar- 

 supials, in Iho latter of which only one (pp. 3) is so 

 changed, and it is consequently a difficult (juestion to say 

 whether these South American forms .should be classed 

 as creodonts or marsupials. 



Be this as it may, it is (juite evident that creodonts 

 and marsupials are nearly related, and have probably 

 both sprung from Mesozoic ancestors. The next ques- 

 tion is whether these Mesozoic ancestors should be 

 called creodonts or (in a wide sense) marsupials. Un- 

 fortunately the degree of preservation of the compara- 

 tively few and imperfect known remains is such as to 

 preclude a definite answer being given to the ques- 

 tion. Dr. Wortman, to whose opinion I attach great 

 value, inclines to the belief that they were marsupials. 

 Personally, basing my opinion on the restricted tooth- 



