Dr. W. D. Matthew—Relationship of the Sparassodonta. 531 
ones lying side, by side. Shear planes are, however, constantly 
forming and breaking up, and displacing the parts of the crystals 
more or less. Figs. 4, 5, and 6 show this clearly. As, however, 
the ice grains are continually changing shape, some growing and 
others decreasing in size, these shear planes in time disappear as they 
have partly done in Fig. 7. 
In the paper already referred to this interchange of molecules, which 
causes some glacier grains to grow and others to decrease, was regarded 
as the cause of glacier motion, and should produce true viscous motion 
in the glacier as a whole, even though it may be unsound to regard an 
ice crystal as truly viscous at right angles to the optic axis. 
EXPLANATION OF PLATES XXIII AND XXIV. 
Fies. 1-2. Full-sized photographs of thin slices of ice, about 0°03 inch thick. 
_,, 98-7. Lines and surfaee-markings obtained by taking rubbmgs of the actual 
surfaces of the glacier ice in ice-caves. They consist of alternate 
ridges and furrows. 
II.—Tuer RELATIONSHIPS OF THE ‘ SPARASSUDONTA.’} 
By Dr. W. D. Marruew, of the American Museum of Natural History, New York. 
[J\HE Sparassodonta are an interesting group of extinct mammals 
found in the Tertiary formations of Patagonia. They appear to 
have taken the place of true Carnivora in South America during most 
of the Tertiary period, as the Carnivorous Marsupials do in the modern 
_ fauna of Australia. Their relationship has been in dispute since they 
were first made known to science by the distinguished South American 
paleontologist Florentino Ameghino. In dental formula and other 
characters of the teeth and jaws they agree with the Carnivorous 
Marsupials, and show a high degree of adaptive specialization for 
_predaceous habits. Dr. Ameghino was able to determine, however, 
from specimens in his collection, that the replacement of the premolars, 
in certain genera at least, was more complete than in modern mar- 
supials, and approached that of the placental carnivora. He regarded 
-them, therefore, as an intermediate group, ancestral to the Creodonta 
and modern Carnivora. Mr. Tomes, in his interesting study of the 
-enamel structure of the Creodonta and allied groups, has recently 
shown that in one genus of Sparassodonts the enamel shows the more 
highly differentiated structure of the true Carnivora, instead of the 
more primitive structure of Marsupial Carnivores. The collections 
made by the Princeton and American Museums in the Santa Cruz 
formation of Patagonia contain a splendid series of Sparassodont 
remains, including skulls and skeletons of the principal genera, which 
have recently been exhaustively studied by Dr. Sinclair. These show 
that the Sparassodonts agree with Marsupials, and especially with 
Thylacinus in almost all the distinctive features of dentition, skull, and 
skeleton characters by which they are separated from the placental 
mammals. Mr. Sinclair gives an extended list of the marsupial 
1 « Marsupials or Creodonts?’’ R. L.in Wature, March 21st, 1907. C.S. Tomes, 
‘““On the Minute Structure of the Teeth of Creodonta, etc.’?: Proc. Zool. Soc. 
London, June, 1906. ‘W. J. Sinclair, ‘‘ Marsupialia of the Santa Cruz Beds” : 
. Reports of the Princeton Expeditions to Patagonia, vol. iv, part 3, Sept., 1906. 
