NOTES. 255 



the Hindoos and the Thibetans. The same writer as- 

 serts, that Cecrops and Pythagoras were acquainted with 

 this system, of Egyptian numeration ; and that it took 

 its origin from the lineary hieroglyphical arithmetic, in 

 which perpendicular strokes have a value of position, 

 while several rows of horizontal bars denote tens, and 

 the multiples of ten (Gatterer, Weltgeschichte bis 

 Cyrus, p. 586). According to this hypothesis, the nota- 

 tion peculiar to the Hindoos would have been introduc- 

 ed for the second time into Europe by the Arabians ; but 

 these assertions do not seem to rest on very solid foun- 

 dations (Kircher, Obel. Pamph., p. 46l). We know, 

 that among the Romans, whose numerical system is in- 

 finitely more imperfect than that of the Greeks, the 

 unit changes its value according as it is placed before 

 or after the signs of five or of ten. A real value of 

 position is found in the notation, which, according to 

 Pappus, Apollonius made use of for the myriads, (De- 

 lambre, Arith. des Grecs dans les Oeuvres d'Archimede, 

 1807, p. 578) : but none of the nations, of which we 

 have authentic accounts, appear to have attained this 

 simple and uniform method, which was followed from 

 remote antiquity by the Hindoos, the Thibetans, and 

 the Chinese. 



Page 128. Twelve Sunas. The inhabitants of Ota- 

 heite divide the year, not into twelve, but into thirteen 

 months, or moons, to which they give the names of the 

 sons of the Sun (Missionary Voyage to the Pacific 

 Ocean, 1799, p- 341 — 344). This division by thirteen 

 is very extraordinary no doubt ; but we know, that peo- 

 ple far advanced in civilization have long stopped in 

 their calendar at numbers the least fitted for the divi- 

 sion of time. See the valuable researches of Mr. Nie- 



