716 ELEPHANTS xxix. 4- 



lines, in which evolution proceeded in a parallel manner. From the 

 end of Miocene times onwards the very elongated jaws began to 

 shorten, and it was presumably at this time that the trunk became 

 long and the typical elephant-like habit was adopted. Certainly not 

 all animals having elephant-like appearance belong to the 'main' 

 elephant line, and the shortening of the lower jaw took place at 

 different times in the various stocks. Thus the members of the 

 Pliocene and Pleistocene genus *Anancus, though extremely like 

 elephants in general shape, had bunomastodont molars. This line 

 retained little cusps between the main ridges, such as were present in 

 the Miocene gomphotheres. *Stegomastodon was a related animal that 

 lived on in South America until as late as a.d. 200. 



The species of * Serridentinus represent another line. Here there 

 were no little cusps on the teeth, but the grinding area was increased 

 by extra 'serrate' ridges. This type retained the long lower tusks into 

 the Pliocene, but then tended to shorten them, though never to the full 

 elephant condition. The line that produced the modern elephants can 

 first be recognized in the Lower Miocene, by the fact that the cusps 

 are united into sharp ridges. This condition presumably marks the 

 transition to feeding on hard grasses, which have to be cut, rather 

 than on softer stalks, which can be ground. Some members of this 

 'zygolophodont' stock retained few ridges, and presumably a browsing 

 diet, even into the Pleistocene, although they acquired short lower 

 jaws like those of the elephants. These animals are therefore 'masto- 

 dons', and it is unfortunate that the rules of priority require that they 

 shall be called *Mammut (= *Zygolophodori). In * Stegolophodon and 

 *Stegodon of the Upper Pliocene and Pleistocene, however, the lower jaw 

 was already short and the arrangement of the skull elephant-like, with 

 huge curved upper tusks. From some such form the two modern ele- 

 phants {Loxodonta and Elephas) and the mammoths (*Mamtnuthus) 

 have been derived, but the details of the evolutionary sequence become 

 here even more involved. We possess such great numbers of teeth, show- 

 ing all sorts of detailed differences, that to arrange them in evolutionary 

 sequences is mostly a matter of guesswork. Osborn preferred not to 

 deal with individual teeth but only with the fifty or so complete skulls 

 that have been found, and these fall into three groups. The mam- 

 moths, * Mammuthus (= * Archidiskodon = *Mammonteus) had very 

 long curved tusks, turned upwards at the tips. There were several 

 'species', including such forms as M. primigenius, the woolly mam- 

 moth, common in Europe during the Pleistocene and surviving until 

 recently in Alaska and Siberia. Many carcasses of these animals have 



