534 
Proceedings of the Royal Society 
We make our pound of seven thousand troy grains, and come down 
again with sixteen ounces to the pound ! 
The introduction of the Indian numerals and notation has brought 
the inconvenience of these haphazard schemes into strong relief. 
The whole power of this algorithm comes from its uniformity. 
The old scheme of counting by help of letters had proceeded 
decimally, its great convenience having led to its use among the 
Arabs, from whom it passed into Greece. In this scheme the value 
of the letters depends on their place in the Hebrew Elif Be, which 
place is fixed in the Arab’s memory by the rhythm “ ebjed hevves 
hota kelmen,” &c., while the Greeks had to supply two new 
characters to fit it to their alphabet. The first group of nine letters 
are taken to signify units, the second group tens, and the third 
group hundreds. But the marks, in the Indian method, rise in 
value ten times for each step on the scale, and thus ten characters 
serve, and much more than serve, for the former twenty-eight. 
We, who have never had to use the older method, can hardly 
appreciate the magnitude of the improvement. Adopted at once by 
men of science, it led to the decimal division of the radius, and to 
the construction of the canon of sines in its modern form. Passing 
to commercial men, it greatly facilitated their computations. In 
every branch of business its influence is felt. Thus Fahrenheit, 
when arranging his thermometers, divided the capacity of the bottle 
decimally, and estimated temperature by the expansion of mercury 
in ten- thousandth parts of its bulk ; while Celsius, proceeding in 
another direction, placed one hundred degrees between the tempera- 
tures of freezing and boiling water. The chemist makes his analyses 
in hundredths ; the banker discounts per cent. \ in every quarter the 
struggle is in favour of decimals. Gunter contrived his chain of one 
hundred links in order that there might be one hundred thousand 
square links in an acre ; the engineer graduates his levelling staff 
not in feet, inches, and eighths, but in feet and hundredths. 
There is no doing without decimals ; when, in making a propor- 
tion, we have to compare twm quantities of one kind, we, as the 
arithmeticians say, bring them both to one denomination : 2 cwt. 3 
qrs. 17 lbs. 11 J oz. must be brought to quarter ounces, of which 
there are 20 845 in this quantity. That is to say, having found 
our old system to be unworkable, we have recourse to counting in 
