xxix. 4 EVOLUTION OF ELEPHANTS 715 



The skull was elongated and small tusks were present, but the denti- 



'2. 1. 'J. "3 



tion was nearly complete, . The molars were bunodont and 



2.0.3.3 



carried only two cross lophs, a condition easily derived from that of 



a condylarth quadritubercular tooth. From some such animal arose 



ultimately so great a collection of types that Osborn in his study of the 



group recognized 350 species, only two living at the present day. 



*Moeritherium survived into the Lower Oligocene, where there is 



found also *Phiomia, about twice as large. Both upper and lower jaws 



of this animal carried tusks and the whole front of the head was 



greatly elongated, with formation of a long diastema. The wear of the 



lower incisors shows that they were used for digging. The molars 



carried three low ridges and were all used together, not successively. 



The similar *Palaeomastodon lived at the same time and was about 



6 ft high. 



After this period there is a gap, probably covering 10 million years 

 or more, in our knowledge of elephant evolution, but it is clear that 

 throughout this long period the stock must have continued with little 

 change, as a race of animals with long digging tusks and rather mobile 

 face and lips, becoming gradually larger and elongating the face more 

 and more to enable the ground to be reached. In the Miocene is found 

 a considerable variety of these long-jawed animals, the various species 

 of the genus *Gomphotherium (= *Trilophodon). The teeth of the 

 earliest of these long-faced elephants were used all at once, not in 

 series; later the premolar teeth tended to be reduced and the molars 

 became covered with an increasing number of cusps, arranged to make 

 a number of cross-ridges, seldom, however, more than five. From the 

 low cusps they are known loosely as 'mastodonts', but the colloquial 

 terms for description of elephants are used in senses almost as varied 

 as the 'scientific' names; as in other branches of knowledge, great 

 abundance of information has led to confusion of terminology. 



From this Miocene stage onwards the study of proboscidean evolu- 

 tion becomes a desperate attempt to sort out huge numbers of fossil 

 specimens (often, however, only molar teeth) into truly phylogenetic 

 lines. Osborn made an heroic effort to recognize only fully docu- 

 mented sequences, but even with the wealth of material available it 

 is rarely possible to say with complete certainty that one type has 

 evolved into another. The interpretations of the sequences and their 

 expression in classificatory terms vary considerably even in the hands 

 of the most careful interpreters of Osborn's work. However, it is 

 probable that in this mass of material there can be seen several distinct 



