Chap.m. ANTIENT METAPHYSICS. 



3»^ 



rians, where it certainly was In moft antient times. From the Ro- 

 man notation of numbers, we may learn the whole progrefs of the 

 art as it was firft pradifed: They began by marking unity by a fimple 

 ftraight perpendicular Kbe: With iuch lines they went on counting tiif 

 they came to the number ^i'^: And this they marked by two of thefe 

 lines joined together af the loWer end, fo as to form an acute angle, 

 and make this figure, V. Then they went on adding to this num- 

 ber, V, four units, marked, as I have laid, by ftraight lines: And thus 

 they arrived at the number tcn-^ which was very properly, 1 think, 

 marked by adding the X.wo Jives together, fo as to fet the one above 

 the other; which makes the figure X. And then they went on count- 

 ing by 'tihits, fives, and tens, till they came to greater numbers, fuch 

 as 50, 100, 5C0, and 1000, which they marked by the letters L, C, 

 D, and M. It appears, therefore, that the firft arithmetic was qui- 

 nary, and not decimal, as among us ; and that men at firft pro- 

 ceeded in numbering by Jives, as Proteus did in numbering his fea 

 calves *. And it was very natural that men, when they firft began 

 to number, fliould firft count the fingers of one hand; and, flopping 

 there, fhould begin again, and go to the fingers of the other hand. 

 In this way they counted 10, and there they alfo ftopped, and again 

 went over the fame numeration: And it was this fecond flop, which, 

 I am perfuaded, introduced the decimal arithmetic. 



And here it may not be improper to obferve, how much better 

 the notation of numbers, by what we call Arabian cyphers, but which^ 

 as is v/ell known, are truly Indian, is than the Roman notation; 

 by the Indian notation all numbers, however great, are exprefled by 

 nine figures, with the addition of zeros, denoting only that any of 

 the nine figures prefixed is to be underftood as multiplied by as 

 many tens as there are zeros. In this way, not onlv the o-reateft 

 numbers are eafily marked, but every operation upon them, of ad- 



R !• ^ ding, 



* Odyff". lib. 4. V. 412. 



