Mr. Tredgold on the Weights and Measures. 303 



be measured, than on any relation to a standard in nature. The 

 French have taken their standard from the globe, and yet 

 have omitted to provide a measure for one ; they have affected 

 to borrow from nature, and have overlooked the most ob- 

 vious of natural distinctions. 



I am glad that so much of our old measures and weights are 

 to be preserved, for I have not yet obtained enough of know- 

 ledge to enable me to do without referring to my predecessors, 

 and I am extremely unwilling to encounter any additional 

 difficulty in holding communication with them ; it makes one 

 think seriously about the advantanges promised by the change 

 to the decimal system. What foundation has this decimal 

 system in the nature of things? — will it continue for ever to be 

 the best possible system of notation ? or, is it itself imperfect 

 and likely to be changed as soon as a better shall appear ? If 

 the latter be the case, what is to be done with a system of 

 weights and measures formed on the decimal scale ? For my 

 own part, I am one of those who think our system of weights 

 and measures to be founded on more rational principles than 

 our notations. If the notation had been formed on principle, 

 surely 10 would never have been fixed upon for the basis of 

 the system ; it has only two factors, and one of these is a 

 prime, which is not so frequently a factor as the prime below 

 it ; and which renders it often more convenient to work by 

 vulgar fractions than by decimals. That is, the prime 3 oc- 

 curs more frequently in calculations than the prime 5 ; and 

 whenever the prime 3 is a factor of division, the decimal no- 

 tation is incapable of expressing the quotient. The decimal 

 system owes all its advantages to the happy thought of ar- 

 ranging numbers according to their powers ; but this arrange- 

 ment is not peculiar to it; in algebra we adopt a modification 

 of the same principle, but there it is not limited to 1 digits, 

 for we are able to arrange powers or any multiples of powers ; 

 and it will be of incalculable advantage to obtain an equally 

 general arrangement for common numbers. Our present sy- 

 stem is very unwieldy when there is occasion to express 

 large numbers, perhaps quite as much so as those abandoned 

 in favour of the Arabic notation were for ordinary numbers ; 

 and I think no one will venture to say that it is impossible to 

 invent a more perfect notation than the one now in use. Ap- 

 parent simplicity is not a test of the merit of any invention, 

 unless that simplicity be accompanied by fitness for the ob- 

 jects it is to accomplish; and it is not much in favour of the 

 decimal scale to remark, that there are 4- out of the 9 digits 

 <>! which the reciprocals cannot be expressed in finite terms; 

 vi/.. >>,, J, I, and \ ; and that to express \ we must employ 



two 



