Progress in Mammalian Palaeontology. — Oshom. 205 
the comparatively recent effort to trace* all rhinoceroses 
through the Oligocene Ace rather ium occidentale Leidy as 
the stem form. 
The polyphyletic law is an outgrowth of four different 
kinds of evidence. First, that the stem forms are very much 
older than we supposed them to be; we placed them in the 
Pliocene and Miocene, they have now been traced to the 
Oligocene and Eocene. Second, as a consequence of this, 
certain modern genera of mammals have their own ancestry, 
apart from that of closely related genera, as far back as the 
Oligocene and perhaps Eocene. The most conspicuaus ex- 
ample of this is the tracing back of the Dholes (genus Cyon) 
among the Canidse, to an Oligocene form, showing that Cyon 
separated from Cants in the Eocene (Wortman and Mat- 
thew) 1 . Third the polyphyletic law is the result of local 
adaptive radiation or divergence apparently of habit either 
by choice or by necessity. For example, among the horses 
it separates off the grazing types (Pmtohippus), which are 
naturally progressive, from the browsing types (Uypo/rip- 
pus), which are naturally conservative, both found in the 
same locality (Fig. 4). It thus splits up animals living in a 
single region into a number of contemporaneous types or 
genera which may coexist throughout long periods; it is a 
segregation, functional rather than adaptive. Fourth, the 
polyphyletic law results from the invasion into a region 
of a generic or specific phylum which has evolved on an- 
other continent ; for example, the Eurasiatic Teleocera 
came in among the American rhinoceroses in the [Middle 
Miocene. 
This polyphyletic law has now been demonstrated 
( ( >sborn 2 ) among the rhinoceroses both of Eurasia and of 
North America, and is the key to the comprehension of this 
group; in Fig. 3 printed herewith it is shown that there are 
not only three families, namely, cursorial (Hyracodontidae), 
aquatic ( Amynodontidae) , and terrestrial (Rhinocerotidae), 
but that the last family splits up into six and possiblv seven 
phyla, many of which are contemporaneous; and the ten- 
1 The Ancestry of Certain Members of the Cattidae, theViverridae, <ui,/ 
Procyonidae Bull. Amer. Mus. Nat. Hist., vol. xii, 1899, pp. 139-148. 
2 Phytogeny of the Rhinoeeroses of Europe. Bull. A.mer. Mus, Nat. 
Hist., vol. xili, 1900, pp. 229-267. 
