( M7 ) 
a doler Examination, and farther Inquiry, I found 
there was no Reafon from any of thefe Dates to lup- 
pofe, it was really true in Fa<ft. For the Helmdon 
Date inftead of 133, fliould, as I then flic wed, be 
read ^133 ; the Colchefler Date 1490, inftead of 
1090 ; and that at JVidgel Hall has no Arabian Fi- 
gures in it, the Characters 1 and 6 not being Num- 
bers, but the initial Letters of two proper Names 
I G, in the ufaal Form of thole Letters in that Age. 
But there has been very lately read before thia 
Society, an Account of a Date at Worcejler , more 
antient than any of the three former ; namely 
or 97V, in which the Unite is a Roman Nu- 
meral, and the other two are taken for Indian Fi- 
gures. I oblerved in my former Paper, that luch Mix- 
tures were fometimes found in antient Numbers; 
tho’ in what Manner they were fo uled, I did not 
then explain, but for Brevity contented myfelf with 
refering to the Algebra of Dr. W allis, a Book lb 
very well known. The DoCtor thought it neceflary 
to take Notice of this, in Order to account for his 
Way of reading the Helmdon Date, in which the 
only is a Roman Numeral. And I had myfelf met 
with a few Inftances of it in Dr. Mead's, Manulcript 
of Boethius, as 00019 and dcc68, where the Hun- 
dreds are numeral Letters, and both the Decimals and 
Unites Arabian Figures a . But ’tis obfervable, this is 
not done promifcuoufly, but the largeft Numbers 
are always Letters, and the lefter, Figures; as in 
the Helmdon Date. And Mabillon has obferved, 
that in a curious manulcript Copy of Thomas a Kem* 
* De Arith.Lib.il. 
S t /ir. 
