THE MAMMALIA OF THE DEEI' ElVEli BEDS. 119 



described in the preceding pages, we find that it is fitted to throw very welcome 

 light upon some disputed questions of evolutionary philosophy. In a former paper 

 (ISTo. 33, p. 371) I considered the problem as to whether the differentiation of any 

 group is a steadily advancing one (or retrograding, as the case may be), interrupted 

 only by relatively stationary periods of rest, or whether it should rather be regarded as 

 progressing in a spiral, advancing, on the whole and in the long run, but with many 

 deviations, setbacks and retrogressions. The evidence then available from fossil 

 mammals did not seem to give any very definite answer to this question, and, while 

 the new material offers important help in the solution of the problem, we cannot hope 

 to solve it definitely. The grand difficulty in the way of applying the results drawn 

 from the study of mammalian phyla to the solution of such general questions lies in 

 the fact that only very rarely can we construct a phylogeny of species as distin- 

 guished from that of genera, and the latter are too vague for the purpose. A hardly 

 less formidable difficulty is caused by the influence of migrations from one region to 

 another. The phenomena of parallelism, interesting as they are in themselves, are 

 often impossible to distinguish from the effects of a common inheritance, and the 

 tendency in successive genera to repeat similar cycles of specific variation only adds 

 to the confusion. Sometimes an apparently simple and easy step in advance is de- 

 layed for an incredibly long period. Thus in the little Mesoliipjius the ulna is as 

 much reduced and as frequently coiissified with the radius as in the very much larger 

 Ancliitherium which appears so much later in time. The lachrymal pit is constant 

 until we reach Miohippus, when it becomes subject to variation, one species at least 

 being devoid of it, while in the much more advanced genera, Protohippus and Hlppar- 

 ioti, the same variation is found, some species having the pit and others not. Yet a 

 phylogenetic scheme founded upon the presence or absence of the lachrymal depres- 

 sion would lead to absurd results. Still another obstacle to progress in these ques- 

 tions is found in the conditions of preservation of the fossils. As we examine large 

 series of forms from several successive horizons, we find that the great majority of 

 species and genera are confined to one or two formations and that each succeeding 

 fauna is recruited partly by migration from other regions, partly by the rapid expan- 

 sion of a comparatively few adaptive and plastic types, while most of the forms 

 which were especially fitted for the older conditions die out under the new. Now, 

 the collections contain, principally, the dominant and abundant species of any given 

 horizon, and these are frequently not the ancestors of the species which will be dom- 

 inant in the succeeding period. Only rarely do we find so many lines keeping on 

 without break from one horizon into another as those which pass from the White 



A. p. s. — VOL. XVIII. P. 



