A Note on the Roman Numerals. 
173 
1907-8.] 
bending down each finger as it is used for enumeration. They begin 
with the little finger. This little finger, then, is one. Now six is five 
(that is, one whole hand) and one more. We can easily see then why 
their word for six comprehends the word for one. . . . Again, when 
they have bent down the eighth finger, the most noticeable feature of 
the hand is that two fingers, that is, a finger and a thumb, remain 
extended. Now the Aht word for eight comprehends “ atlah,” the word 
for two. . . . Their word for nine comprehends “ tsowwauk,” the word 
for one. Nine is ten (or two whole hands) wanting one * ” (p. 337). 
“ The Zamuca and Muysca Indians [Humboldt, Personal Re- 
searches, ii. 117] have a cumbrous but interesting system of numeration. 
For five they say, ‘hand finished. 5 For six, ‘one of the other hand, 5 
. . . for ten they say, ‘ two hands finished, 5 . . . twenty is the feet 
finished ; or in other words ‘ man, 5 because a man has ten fingers 
and ten toes, thus making twenty 55 (p. 337). 
“ Speaking of the Guiana natives, Mr Brett [Brett, Indian Tribes 
of Guiana , p. 417] observes . . . ‘forty-five is laboriously expressed 
by ... “ two men and one hand upon it 55 5 55 (p. 338). 
That the Romans, too, had some method of counting by the fingers is 
shown by their use of the word digiti (“ digits, 55 as we still term the lower 
numbers), in such phrases as : “ in digitos rem redire,” “ in digitos mittere,” 
“ in digitis constituere,” “ digitos tollere.” Vitruvius ( ioc . cit.) refers to the 
body as the source or origin of numbers : “ Si autem in utrisque palmis ex 
articulis ab natura decern sunt perfecti,” and “ Ergo si convenit ex articulis 
hominis numerum inventum esse”; and Ovid {Fasti, iii. 121-3) has this 
allusion to what appears to have been the contemporary practice of digital 
numeration : 
“ Annus erat, decimum cum luna receperat orbem. 
Hie numerus magno tunc in bonore fuit. 
Seu quia tot digiti, per quos numerare solemus.” 
And Juvenal {Satires, x. 240-1) implies that the system was not restricted 
to the lower numbers : 
“ Felix nimirum, qui tot per saecula mortem 
Distulit atque suos jam dextra computat annos.” 
It is also evident that the first ten numerals admit of being interpreted 
and explained as successive steps in a digital numeration, though their 
forms have been conventionalised and even assimilated to alphabetic 
characters : 
A.— I, II, III, IIII, V. 
b.— vi, vii, viii, vim, x. 
