The distribution of mankind is therefore only a distribution into 'varieties', which are commonly designated by the somewhat indefinite term 'races'. As in the vegetable kingdom, and in the natural history of birds and fishes, a classification into many small families is based on a surer foundation than p 356 where large sections are separated into a few but large divisions; so it also appears to me, that in the determination of races a preference should be given to the establishment of small families of nations. Whether we adopt the old classification of my master, Blumenbach, and admit 'five' races (the Caucasian, Mongolian, American, Ethiopian, and Malayan), or that of Prichard, into 'seven races'* (the Iranian, Turanian, American, Hottentots and Bushmen, Negroes, Papuas, and Alfourons), we fail to recognize any typical sharpness of definition, or any general or well-established principle in the division of these groups. [footnote] *Prichard, op. cit., vol. i., p. 247.
Citable passage
Cosmos
Alexander von Humboldt. Cosmos. 1845.
Citation
Alexander von Humboldt. Cosmos. 1845. Premodern Concordance passage passage-7d27c9b4ee7ef5d352e2.