investigating the early collections made in Japan. Modern work in China, and especi* 

 ally central China, has given overwhelming evidence and established beyond question 

 Asa Gray's conclusions. There are many instances in which only two species of a genus 

 are known — one in the eastern United States and the other in China. Noteworthy 

 examples are the Tulip tree, Kentucky Coffee tree, the Sassafras, and the Lotus Lily 

 (Nelumbium). A considerable number of families are common to both countries, and 

 in most instances China is the dominant partner. Usually the U. S. A. have one and 

 China several species of the same genus, but here and there the opposite obtains. 

 Magnolias afford a good illustration of this affinity. This genus, absent from Europe 

 and western North America, is represented by 7 species on the Atlantic side of the 

 North American continent, and by 19 species in China and Japan." 



Then follows a list of 15 genera common to China, Japan, and the Atlantic side 

 of the United States, but not now living in Europe. Of the 15 genera 4 have already 

 been found in the Reuverian, viz. Magnolia, Stewartia, Berchemia, and Nyssa, as also 2 of 

 the 4 previously mentioned, the Tulip tree (Liriodendrori) and Nelumbium. 



Comparison has been made between the Reuverian flora and the flora of the 

 Yangtse Valley, because the latter is more closely allied to the Reuverian than is any 

 other, and also because it serves particularly well to illustrate the effect of high mountains 

 upon the preservation of a flora under an oscillating climate. A reference to the distri* 

 bution of species shows, however, that there is also a considerable alliance between the 

 Reuverian flora and the flora of the Himalaya. A close relationship has been proved 

 to exist between the flora of Western China and the Himalaya (Dr. Augustine Henry 

 and others) ; but the distribution of species in these regions is as yet only partially worked 

 out, and until more is known of the living floras, we shall not be in a position to solve 

 this problem. Still, we may suggest three factors which singly or combined may account 

 for the relationship between the Reuverian, the Himalayan, and the Chinese floras. 



(1) It has been suggested (Wilson) that the Himalayan flora is an extension 

 westward of the Chinese. If this is so the extension would date back to a time, pro= 

 bably Pliocene, anterior to the cutting of the great gorges. In that case the flora must 

 have entered these regions as an upland flora, and, as the climate cooled, it must have 

 descended. 



(2) It is both possible and likely that in part the Himalayan flora may be a 

 return wave from that part of the migrating flora which passed coastswise into the 

 Malay Peninsula, when the other stream entered the Yangtse Valley. Whether the great 

 alternations of climate in the north extended into tropical latitudes we do not know; 

 but it is possible, for there are clear signs of a former great extension of the Himalayan 

 glaciers. An interval of interglacial warmth might therefore cause a return migration 

 to cooler regions, the path being open, by way of the Mekong, Salwen, Irawadi, 

 Brahmaputra, etc. 



(3) Yet a third suggestion may help to account for the resemblance between 

 these floras. The immense area of high land in Central Asia, including Tibet, the 



22 



