MISCELLANEOUS ESSAYS. 345
did ufe their letters for numerical the numbers which letters were firft
charatUrs ; but from the manner ufed to exprefs ? And what reafon
in which they ufed them, I draw can be afiigned why D, the firffc
my firft argument, to prove that letter in the Latin worA decevi, iex\j
letters were nc)t thus ufed by the (hould not rather have been chofen
Romans, Every letter in the alpha- to ftand for that number than for
bet was ufed to denote feme num- 500, becaufe it had a rude rchm-
ber by the Creeks and Orientals, blance to half an Ml But if thefe
and each letter denoted a lefs or qaeflions could be fatisfactorily an-
gieater number, as it was nearer or fwered, there are other numeral
more renotc from the firil letter in letters, which have never yet been
their alphabetical order, and no accounted fox at all. I think thefe
letter, which in the order of the confiderntions render it probable,
alphabet Itands after another, ever that the Romans did not, in their
denoted a number lefs than the let- original intention, ufe letters to
ter thai Hands before it. Now, if exprefs nujubers at all ; the moll
the Romans, who derived their natural account of the matter feems
letters originally from the Greeks, to be this :
had derived alfo their numeration The Romans probably put down by letters, it is in the highcll de- a fingle Itroke I for one, as is ftiil gree probable, that thefe particu- the praftice of thofe who fcore en lars would have been the fame in a fiate, or with chalk ; this ftroke both: but as not one third of the I they doubled, trebled, and qua- Roman letters are numerah, fo nei- drupled, to exprefs 2, 3, and 4.: ther is the numeral value of thofe thus, II. lU. IV. So far they that are fo, more or lefs, accord- cculd eaiiiy number the minums, ing to their place in alphabetic or flrokes, with a glance of the order ; for D and C, which are eye, but they prefently found, that among the firft letters of the alpha- if more were added, it would fooji bet, and M and L, which are in be neceflary to tell the ftrokes one the middle, are of much greater by one: for this reafon, when they numeral value than X and V, came to 5, they expre'fed it by which are near the end. joining two flrokes together in an But it has been fuppofeJ that the acute angle, thus V, which will Romans ufed M to denote 1000, appear the more probable, if it be becaule it is the firft letter of M/V/i?, confidered, that the progreiiion of which is Latin for looc ; and C to the Roman numbers is from 5 to 5, deno'e 100. becaufe it is the firll /. e. from the fingers on one hand letter of Centum, which is Latin for to the fingers on the other. ICO. Your correfpondent alfo fup- Ovid has touched upon the ori- pofes, that D being formed by ginai of this in his Fajiorum, lib. dividing the old M in the middle, iii. and Vltruv. lib. c. I. has made was therefore appointed to ftand for the fame rem.ark. 500, that is, half as much as the After they had made this acute M flood for when it was whole ; and angle V. for five, they added fingle that L being half a C, was, for the ftrokes to it to the number of 4, fame reafon, ufed to denominate thus, VI. VIL VIII. VIIIL and 50. But what reafon is there to then as the minums could not be fiippofe, that 1000 and 100 were further multiplied without confu-
fion.