108 On a Proposed Neiv Method of Weighing, 



scale : then we have 4 4 grains which, deducted from 9*9 

 grains = 5*5 grains, and '0043 — "0002 = "0041 indicated by 

 the rider. The fineness thus indicated is obviously -5541 : 

 or a small card of reference might be used for showing at a 

 glance the complementary value of any combination of grain 

 weights and tenths of a grain in these extreme cases. 



I will now endeavour to enumerate the special advantages 

 of this method, premising that weights as much in use as 

 those of the bullion assay er, even when lifted with ivory or 

 horn forceps, are liable to wear, and that wear of the weights 

 is a dangerous source of error. 



1. The proposed method is in all respects equivalent to 

 double weighing ; it is independent of the relative lengths 

 of the arms of the beam, whether arising from original 

 imperfect workmanship or from permanent molecular 

 alteration, or from daily vicissitude. 



2. The weighments are all made with an uniform load, 

 which satisfies another condition of accuracy more exactly 

 than the common method, the sensitiveness of the beam 

 varying with the load. 



3. Substituting the proposed new series of weights for the 

 old (for those described on page 103) we devote the latter 

 henceforth to the purposes of reference, upon which aplication 

 a few sentences may now be bestowed : — 



With the rider over end knife-edge, each weight of the 

 new series added to that bearing the same mark in the old 

 series should effect an equipoise, thus :■■ — 



er =10 grains. 



Mark. 



•99 9-9 grains ~\- WESM -\- '1 grain =10 grains. 



•98 



•97 



and so on. And for further verification of the new weights, 

 interchanges, too obvious to require explanation, can be 

 employed. Now, supposing a weight of the new series to have 

 worn in use, say, for example, that the weight marked '97 has 

 become a thousandth of a grain short of its real weight (*2 of 

 a grain) when just ; this becomes evident on placing it in the 

 pan with the "97 reference weight, the rider also in the pan 

 or over the end knife-edge ; this combination of weights will 

 fall short of an exact equipoise by '001 of a grain. To make 

 a fresh weight in place of the defective one, we have only to 

 fashion and reduce a piece of platinum or gold until with 

 -the old *97 weight and the rider it effects an equipoise. 



Old Series 



+ New Series -J- 



Eider = 



10 



Eeal weight. 



Eeal weight. 







9 "9 grains 



+ o + 



•1 grain = 



10 



9-8 „ 



+ -i -h 



•1 ,, = 



10 



9-7 „ 



+ -2 + 



•1 „ = 



10 



