A study of the balance

This research was not undertaken with the deliberate aim of exceeding the precision of mass-determinations as carried out in the great institutions specially devoted to this subject. I should have considered it both presump­tuous and hopeless to attempt such a feat, and should probably have aban­doned my studies at an early stage if it had not fortunately Happened that the decisive part of my work had been completed before I came across comparable results obtained elsewhere. It was only then that I realised that the methods evolved had led to a very considerable diminution of the residual errors of weighings, in spite of the inexpensive instrument employed and of the primitive conditions under which the observations were made. A projected series of experiments called for weighings of small masses, up to perhaps 50 grm., within 0.1 mgrm. As I did not possess a balance, I bought the least expensive type of analytical balance, one by Sartorius, to carry 200 grammes in each pan, with a triangular phosphor-bronze beam, 14 cm. long, with agate knife-edges and plane bearings, with a complete arrestment for beam, suspensions, and pans, and with a rider-scale divided to read directly to 0.2 mgrm. by the usual centigramme-rider. The cost at the time (1905) was £6. The balance was described in the catalogue as “sensitive to 0.2 mgrm.,” but I trusted that the required accuracy within 0.1 mgrm. would be obtainable by using the more refined weighing methods of the physical laboratory.