938 



SCIENCE. 



[N. S. Vol. XII. No. 312. 



that the horizon could not be so ancient 

 and that it much more probably belonged 

 to the Oligocene or Miocene. Fortunately, 

 the solution of this problem does not de- 

 pend upon theoretical views of mammalian 

 morphology, but may be decided by more 

 direct and satisfactory evidence. The 

 Santa Cruz beds overlie and are, to some 

 extent, interstratified with the marine Pat- 

 agonian formation, which is very richly fos- 

 siliferous and from which Mr. Hatcher gath- 

 ered very extensive collections of marine 

 invertebrates. These have been carefully 

 studied by Dr. Ortmann, who concludes 

 from his examination that the Patagonian 

 beds are Lower Miocene, and the Santa Cruz 

 beds, as a whole, are, therefore, still younger, 

 that is, Middle, or possibly even Upper Mio- 

 cene. 



The first impression that an examination 

 of a representative series of Santa Cruz 

 fossils makes upon the northern observer is 

 that of strangeness, of unlikeness to every- 

 thing with which his pi-evious studies have 

 made him familiar, and this impression is 

 only deepened as these mammals are com- 

 pared with those of North America and 

 Europe. Disregarding some very doubtful 

 and incompletely known groups, the Santa 

 Cruz fauna is composed of the following 

 elements. 



1. Marsupialia. 



3. Ungulata, 



(a) Typotheria, 

 2. Unguiculata, (6) Toxodontia, 



(a) Insectivora, (c) Astrapotheria, 



(5) Edentata, [d) Litopterna. 



(c) Rodentia. 4. Primates. 



One of the most striking features of this 

 fauna may be described negatively, by what 

 it lacks ; it has no Carnivora or Creodonta, 

 no Chieroptera (though little importance can 

 be attached to this fact, which may well be 

 accidental) ; no Artiodactyla,Perissodactyla, 

 Proboseidea or Hyracoidea. Of the nine 

 orders only four are found in the Miocene 

 of the northern hemisphere (for the Old 



World Edentates, so called, are of no sig- 

 nificance in this connection), and even these 

 common orders are represented by totally 

 different suborders and families. That 

 Patagonia had long been cut ofiP from any 

 land communication with North America 

 seems abundantly clear, i 



The Santa Cruz Marsupials are of two 

 types : (1) Carnivorous animals, which took 

 the place of the carnivores and creodonts of 

 the North ; these find their nearest ana- 

 logues in the Dasyuiidoe of Australia, but 

 there are such important dififerences of 

 structure as to indicate a long geographical 

 separation from that family. (2) Herbiv- 

 orous animals, of small size, quite remote 

 from any of the Australian forms and typi- 

 fied by the existing South American Cceno- 

 lestes of Thomas. 



The Insectivora are represented, so far 

 as is yet known, bj'^ only a single genus, 

 Necrolestes, which, as Ameghino has sug- 

 gested, is very like the ' Cape Golden 

 Moles ' of Africa, a most interesting fact, 

 the full significance of which is not yet ap- 

 parent. 



The Edentata are found in surprising 

 variety and numbers, making up one of the 

 most conspicuous and characteristic ele- 

 ments of the fauna. Forerunners of the 

 huge Ground-sloths (Gravigrada) of the 

 Pleistocene, are extremely abundanb and 

 are represented in the collection by a num- 

 ber of such well preserved skeletons that 

 a comparison with their great descendants 

 cannot fail to yield results of much interest. 

 One difference is obvious at the first glance, 

 namely, the very much smaller size of the 

 older genera. Much the same statement 

 is true of the Glyptodonts, which are very 

 numerously represented by species much 

 smaller and more primitive than their 

 Pleistocene successors. The Armadillos are 

 likewise extremely varied and abundant, 

 representing not only the various modern 

 subdivisions of the family, but also some 



