368 ANNUAL RECORD OF SCIENCE AND INDUSTRY. 



the transactions of the Copenhagen Academy. The conclu- 

 sions in regard to geographical distribution harmonize with 

 the generally received law that the more the classification 

 of a family rests on characteristics which indicate a real re- 

 lationship, the clearer it appears that each subdivision has 

 its own centre of distribution ; and, further, that the greater 

 the differences of organization between the subdivisions, the 

 greater the geographical distances between these centres. 

 Thus the chestnuts, oaks, and beeches, constituting the three 

 groups of this family, afford three principal centres of distri- 

 bution, and cover three large, widely separated geographical 

 regions ; the chestnuts having their centre in the Malay Isl- 

 ands, the oaks in Mexico, and the beeches in South America. 

 The chestnut group, which is sharply separated from the 

 other groups, also has its own peculiar, tolerably well-defined 

 region, manifesting its greatest diversity of form and its 

 purest types on the Malay Islands, es])ecially on Java and 

 Sumatra, where its proper centre lies. But one species pass- 

 es its boundaries toward the west, and plays an important 

 part around the Mediterranean, and three species are found 

 in America; while the typical genus is found exclusively on 

 the Malay Islands, where only a few species of chestnut-oak 

 are found, and not a single true oak. In a similar manner 

 the oak group occurs chiefly in America, north of the equator, 

 forming a second centre of distribution in the mountains of 

 Mexico, where it manifests not only more numerous species, 

 but also greater diversity of organization, than any where 

 else, several lar^e subdivisions beino; found that are not met 

 with elsewhere, while chestnuts and beeches are entirely 

 wanting. Although the beech group exhibits such a pre- 

 ponderance of species in Chile that that country must be re- 

 garded as its proper home, still the species are so scattered 

 that it is difficult, with the present distribution of land and 

 water, to refer all to a single centre. This difficulty is not so 

 great, however, in regard to Nothofagas, which occurs in 

 New Zealand and Van Dieman's Land, since there are other 

 grounds for assuming that these were at one period connect- 

 ed w r ith Chile. Still it seems impossible to refer the species 

 of Fagus %o the same centre, since the nearest related species 

 is separated by 70 of latitude from the beeches of the south ; 

 so that, paradoxical as it may appear, Japan seems once to 



