124 



KNOWLEDGE. 



[June, 1903. 



been carried to anything like the same degree as in the 

 horse tribe ; and has been to a great extent restricted to 

 an increasing complexity in the stiiicture of the cheek- 

 teeth, coupli'd, in the second of the two gronps, with a 

 reduction in the number of tlie anterior teeth and the 

 development of one or two horns in the middle line of the 

 skull. Nevertheless, in both groups American palseonto- 

 logists have been able to trace a more or less complete line 

 of descent. The tapirs, for instance, appear to have 

 originated from a small Eocene ungulate known as 

 systemodon, which comes so close to the early ancestors 

 of the horse, that it was at one time included in the same 

 series. Modern rhinoceroses, in whieli, as we have said, 

 the number of front-teeth is more or less reduced, can 

 likewise be traced back to early Tertiary ancestors without 

 horns, but with a full series of teeth. 



Among the even-toed imgulates, or those in which the 

 toes ai-e symmetrically arranged on each side of a line 

 running between the third and fourth of the typical series 

 of five, one of the most complete lines of descent has been 

 traced out in the camel tribe, now represented only by 

 the tme camels of the Old World, and the llamas of 

 South America. The evolution of this group seems to have 

 taken place in North America in the middle Tertiary period, 

 in the strata of which we meet with an ancestral form in 

 poebrotherium — an animal of the size and proportions of 

 a small gazelle, but with all the essential characteristics of 

 a camel, although with a fuller series of teeth, and feet in 

 which the main bones are not fvised into a "cannon-bone " 

 and traces of the lateral digits still remain. Earlier 

 members of the group show still more primitive characters, 

 and it is probable, although not quite certain, that these 

 in their turn pass into yet more generalised animals, with 

 four complete toes to the feet, and cheek-teeth in which the 

 crowns are surmounted by simple low cusps instead of 

 forming the tall crescents found in camels and llamas. 



Precisely similar differences, both as regards foot- 

 structure and tooth-structure, distinguish the true rumi- 

 nants (deer, antelopes, etc.) from the jjig-like animals. 

 And yet a complete gradation from the former to the 

 ancestors of the latter can be found, although it would 

 perhaps be difficult to construct the actual pedigree. In 

 like manner from the tallest and largest-antlered members 

 of the deer tribe a gradual passage can be traced to the 

 small and hornless dremotherium of the middle Tertiary, 

 from which it is but a step to the still smaller and more 

 primitive gelocus - not improbably the common ancestor 

 of both deer and ehevrotains. Giraffes, again, partly with 

 the assistance of that marvellous creature the okapi, are 

 now known to be connected with ruminants of a much 

 more ordinary type. 



Till quite recently the elephants and mastodons, forming, 

 with the imperfectly known dinothere, the proboseidea of 

 zoologists, constituted an altogether isolated group whose 

 relationship with other ungulates was quite obscure. The 

 wonderful discoveries of Messrs. Andrews and Beadnell in 

 the lower Tertiary deposits of the Egyptian desert have, 

 however, brought to light the remains of two remarkable 

 animals which, as regards both bodily size and the structure, 

 number and mode of succession of their teeth, serve to 

 connect the modern proboscide.ans very closely indeed 

 with the more generalised extinct ungulates. Nor is this 

 by any means all, for there appear to be indications of 

 affinity between these primitive Egyptian proboscideans 

 and the ancestors of the modern sea-cows (dugongs and 

 manatis). This is, however, a subject on which little can 

 be said at present, for the evidence is not yet published, 

 and the writer is consequently not at liberty to mention 

 even such facts as he knows. When fuller investigations 

 have been made into African palaeontology it is probable 



that not only will the puzzle of the origin of the pro- 

 boseidea be completely solved, but their ancestral relation 

 to the sea-cows will likewise become an established fact. 



Reverting once more to the carnivora, it will Ix; remem- 

 bered that the ancestral types of the modern forms are to 

 be found in the extinct creodonts ; and it is probable that 

 from the latter a direct line of descent can be traced 

 through various extinct forms to the civets and mongooses 

 of the jircsfut day. Till the above-mentioned discovery of 

 mammalian remains in the French phosphorites no carni- 

 vora were considered more widely sundered than the civets 

 and mongooses on the one hand and the weasels and their 

 allies on the other ; the structure of the hinder part of 

 the base of the skull being . very different in the two. 

 Now, however, we are acquainted with quite a number of 

 small mamm;ils from the phosphorites which completelv 

 liridge thegap ijetween the grouj)s in question. Since the base 

 of tlie skull has the same general structure in both, the bears 

 were long regarded as near relatives of the weasel trilje, so 

 that both were classified in a single group, to the exclusion 

 of civets and dogs. Palfeontology has, however, revealed 

 the existence of an absolutely complete transition, both as 

 regards dental and foot characters, from dogs to bears. 

 Consequently, not only are we compelled to regard the 

 latter as the descendants of primitive dog-Uke animals, but 

 the similarity of their skulls to those of the weasel tribe 

 must be looked upon as an instance of parallelism in 

 development and not as an indication of genetic affinity. 



Possibly the chain of relationships does not stop even 

 with the affinity of bears to dogs, for the former appear 

 to be intimately related to the sea-bears and sea-lions, 

 although, unfortunately, there is no palaeontological evidence 

 at present available to ccmfirm or disprove the presumed 

 relationship. 



If there be a real connection between the bear-like land 

 carnivores and the sea-bears, it is almost certain that the 

 typical earless seals must have taken origin frtnn a totallv 

 different group of carnivores. And it has been suggested 

 that the typical earless seals may be derived directly from 

 creodonts related to the form described under the name of 

 patriofelis. Since, however, we cannot trace seals Itirther 

 back than the later portion of the Tertiary period, 

 additional palaeontological evidence must lie awaited 

 before anything definite can be said with regard to their 

 ancestry. 



Neither is the ancestry of whales by any means defiuitelv 

 known, although there are some indications of the line of 

 descent of the group represented by the sperm whale and 

 dolphins. The late Sir W. H. Flower was indeed disposed 

 to regard whales as the descendants of early ht)ofed mam- 

 mals, but there was no definite evidence in support of his 

 contention, and the probability is that they are related to 

 the early carnivores. In modern whales all the teeth have 

 the same simple conical form, but in the middle Tertiaries 

 occur the shark-toothed dolphins, or squalodons. in which 

 the front teeth differ markedly in character from those of 

 the cheek series, the latter being serrated and implanted 

 by double roots. Still more generalii^ed is the earlier and 

 much larger zeuglodon. in which the cheek-teeth are not 

 unlike those of certain seals. So far as it goes, then, the 

 palaeontological evidence tends to connect the modern 

 toothed whales with the carnivora, and it certainlv traces 

 their descent from animals departing much less widely 

 from the more ordinary mammalian type. It may be 

 added that certain living porpoises show a few rows of 

 bony tubercles in the neighbourhood of the back-fin, 

 similar tubercles occurring in a more developed condition 

 in some of their extinct relatives. Now with the remains 

 of zeuglodon are frequently discovered a number of bony 

 plates, which there is accordingly strong reason to believe 



