[CAMPBELL] ORIGIN OF THE HAIDAHS 96 



cerning which Latham says : ''In each of these vocabularies Malay words 

 form the greater proportion. In each of them, however, are also found 

 Australian vocables." Sumbawa, to which Timbora belongs, is in the 

 very heart of the Malay archipelago, and most of its numerals are Malay 

 in charactei". Those that are not accord with the numerals of the Haidahs. 

 I am aware that there are some comparative philologists who regard the 

 common possession of a numeral system as one of the most convincing 

 proofs of a common origin. This is a great mistake. The original Celtic 

 numerals have been replaced by the Latin. The Arabic have ditfused 

 themselves in Africa, the Sanscrit in India, and the Malay in Polynesia. 

 In the intercourse of half-civilized or savage peoples with their superiors, 

 no words are more easily lost. Whether the numerals of the Haidahs 

 represent those of ancient Melanesian speech or not, they are an evidence 

 that Malay influences were not sutRciently strong to impose upon them 

 its arithmetical system. 



Of more importance than these are particles, such as the postposi- 

 tions, of which a list of twenty-six is appended. These are Australian as 

 well as Haidah, and, were we in possession of lists of similar parts of 

 speech from Sumbawa and its vicinity, links might be found to unite the 

 far distant vocabularies. The same is true of pronouns, of which, un- 

 fortunately, my collection is small. Nevertheless it will be found to 

 exhibit traces of kinship between the compared languages such as to 

 render complete the cumulative argument for their original unity. 1 

 have before me Adelung's ■' Mithridates," Klaproth's " Asia Polyglotta," 

 Hunter's -'Non-Aryan Languages of India and High Asia," the "San 

 Kokf Tsou Ran To Sets." and many more recent collections of A.siatic 

 vocabularies, in which I have searched in vain for such traces of linguistic 

 affiliation as I have found between the Haidah and the Melanesian of the 

 Malay-Polynesian area. That the resemblance is fortuitous is an impos- 

 sibility to amy one who has made an exhaustive study of languages, how- 

 ever improbable it may appear at first sight to link Australians and 

 Queen Charlotte islanders as members of the same family. 



There is linguistic evidence of no mean order that many American 

 families of man came to this continent by way of the Pacific islands. 

 Such are the Mbaya-Abipones of the Gran Chaco ; the Tupi-G-uaranis of 

 Brazil ; the Caribs, the Huastec-Maya-Qvaiches of Central America, and 

 the Algonquins of the north. All of these are of Malay-Polynesian origin 

 except the Tupi- Guaranis, and they are Melanesian, like the Haidahs. 

 The difficulty of a comparatively savage people traversing a wide ocean 

 is an argument that should not weigh against the demonstration of 

 language. The people of Easter Island came within eighteen hundred 

 miles of the American coast, but, supposing them to have started from 

 the Philijjpines, their route was one of eight thousand miles. Even re- 

 cently, in comparatively small canoes, the islanders of the South Seas 



Sec. II., 1897. 4. 



