96 On a Proposed Ne%v Method of Weighing, 



former in some degree, so as to secure a speed suitable to 

 the quantity of work which is demanded in a given time. 



Chemical balances of the best construction are known to 

 indicate the millionth part of the load, that is to say, with 

 1000 grains in each pan the index will move over one 

 division of the scale for -001 of a grain added. With the 

 assay balance, with 10 grains in each pan, if '001 of a grain 

 added is represented by one division of the scale, the sensi- 

 tiveness falls far short of that of the larger chemical balance ; 

 instead of one-millionth, the quantity indicated is one ten- 

 thousandth of the load ; it is one hundred times less sensitive 

 than the former, although in each case "001 of a grain is 

 indicated by one division of the scale. The oscillations, 

 however, are quicker in the case of the assay balance. 



Each individual assay balance will have its own degree of 

 sensitiveness when adjusted for the slowest speed, or its own 

 lesser degree of sensitiveness when adjusted for a given 

 rapid working speed ; and by lowering and heightening its 

 centre of gravity, we may attain any combination of speed 

 and sensitiveness, gaining in the one while we lose in the 

 other, between these limits ; but we are with each particular 

 assay balance, bound within these limits of performance of 

 the special instrument. Thus, if we have two assay beams, 

 without radical defects, but of different make and weight, it 

 will be found that their performances will differ ; one will 

 prove inferior to the other. If the inferior beam be adjusted 

 to the same degree of sensitiveness as the better constructed 

 beam, it will take a longer time for its oscillations ; or if it be 

 adjusted to oscillate in the same time as the superior beam, 

 it will then fall short in sensitiveness. 



But the knife-edges of the best balances, whether made of 

 steel or agate, wear and become blunted during use, and the 

 performance is, from this particular cause, gradually deterio- 

 rated. A large part of the keeping the balances in repair 

 consists in the reparation, from time to time, of the knife- 

 edges, and the readjustment of them in position in the beam. 

 But between the time when the knife-edges are sensibly 

 perfect and when they require the instrument maker's 

 attention there is of course a small progressive wear and 

 alteration of them, with consequent deterioration of the 

 performance of the beam ; the value of each degree of the 

 index scale becoming from time to time a larger weight. 

 These small progressive differences, within certain limits, 

 may be obviated by adjustment of the ball ; but when we 



