Nov. 28, 1889] 



NATURE 



83 



subcylindrical column totally unconnected with the anterior 

 crescent {c) ; in Hippidiian this pillar retains almost the 

 same form as '\n Hipparion^hvLi becomes connected with the 

 crescent ; while in the existing horses the same pillar has 

 become greatly elongated in anantero-posterior direction. 

 Further, in Hippidium the first premolar, which in modern 

 horses is generally absent, and if present is minute and 

 deciduous, is of very large size, and always persists. 



The Pliocene Eqitits stenonis of Europe forms, however, 

 a connecting link in respect of dental characters between 

 the American Hippiditiin and the modern horses ; and it is 

 therefore to a great extent a matter of individual opinion 

 whether or no the retention of Hippidium as a distinct 

 genus is convenient. A new species referred \o Hippidiiivi 

 is described from Tarija, in Bolivia. Of more typical 

 horses the author describes additional remains of Equns 

 curvidens, E. argentini/s, and E. andium j and he adds 

 to his description a useful word of warning in regard to 

 the many forms of fossil horses from other parts of South 

 America which have been described as distinct species, 

 suggesting that all or several of these may be based merely 

 on individual variations. 



In the second section of the volume we have a descrip- 

 tion of remains of other mammals from the Pampean 



deposits recently acquired by the Museum at Buenos 

 Ayres. The first of these additions is an entire skull of 

 Megatherium americanu7n,\i\\\c\\. shows that our previous 

 knowledge was incomplete. This skull formed part of a 

 nearly entire skeleton of a very large individual found in 

 August 1888 on the Rio Salado, but which is as yet but 

 partially disinterred. It shows that instead of the aperture 

 of the nares being bounded superiorly merely by short 

 nasal bones which did not reach within a long distance of 

 the premaxillae, there was a large prenasal bone extending 

 nearly as far as this point ; while there was also a lateral 

 process projecting forward from the upper part of the 

 maxilla into the nasal aperture. This prenasal bone is 

 4i inches in length, and it is considered probable that it 

 became united with the nasals in the adult. Still more 

 remarkable, however, is the presence of another ossifica- 

 tion extending upwards and backwards from the superior 

 surface of the extremity of the premaxillae towards the 

 prenasal bone, from which it is only separated by a short 

 interval. These two ossifications, we may observe, are 

 evidently a rudiment of the complete bony arch connecting 

 the premaxillae with the nasals in Mylodon darwini, 

 which was on that account generically separated by 

 Reinhardt as Grypotherium ; and they serve to support 



Fig. 2. — The third left upper true moiar of Mastodon hnmboldti; from the Pampean of Buenos A3 res. Two-thirds natural size. 



that the last-named species is not 

 genus in which it was originally 



Prof. Flower's view 

 separable from the 

 placed. 



The author next proceeds to the consideration of 

 the skull of that species of Mastodon which he terms 

 M. antiiim. No mention is made of the earlier name 

 M. cordilterum, which appears to be the proper one for 

 this species ; and in amending the usual spelling M. 

 andium to M. a7itium, one cannot help wondering why 

 the same course was not adopted in the case of Equus 

 andiuift. The object of this part of the work is to show 

 that the reference by the late Dr. Falconer to M. cordil- 

 lernm (as we will call it) of mandibles from Texas, fur- 

 nished with long tusks is incorrect, and that this species 

 really had, like its near ally M. humboldii, a mandibular 

 symphysis of the same general type as that of the 

 elephants, without any tusks at all in the adult. Figures 

 are given of an immature and of an adult skull with the 

 mandible in situ to support this redetermination. Dr. 

 Burmeister then proceeds to institute a comparison 

 between M. cordillcrum and M. humboldti, in which he 

 states that, although very similar, a careful examination 

 shows very clearly the distinctness of the two forms. 

 Here we may observe that it is to be regretted that no 

 comment or reference is made to the notices and figures 

 published by Falconer and other English writers in refer- 



ence to these forms ;] but perhaps the real explanation of 

 this omission is that the libraries at Buenos Ayres are not 

 so well stocked as those of London. According to our 

 author, M. cordillerum is the smaller of the two species ; 

 the length of the mandible from the condyle to the 

 symphysis being 75 centimetres against 85 centimetres in 

 M. hnmboldti J the last dimension agreeing with the 

 British Museum ckull of that species originally described 

 by Falconer in M. andium. Falconer's observations as 

 to the more complicated structure of the molars of M. 

 humbolti are in the main confirmed. A small specimen 

 of a last upper molar referred to this species in the British 

 Museum is (with the permission of Dr. Woodward) 

 figured in the accompanying woodcut, to show the com- 

 plexity of the crown, in which the valleys are much 

 blocked by accessory tubercles. In the early stage of 

 wear of this specimen imperfect trefoils of dentine are 

 shown only on the inner columns ; but when more worn 

 trefoils would evidently also appear on the outer columns. 

 In the well-worn upper molar of M. cordillerum, repre- 

 sented in Plate x., Fig. 5, of the work before us, the 

 absence of a distinct trefoil on the outer columns, which 

 Falconer mentioned as one of the distinctive features of 

 this species, is well shown. Dr. Burmeister further 

 observes that the molars of M. cordillerum are charac- 

 terized by their blackish enamel, and the brown or 



