On a Proposed New Method of Weighing. 93 



Art. XXII. — On a Proposed New Method of Weighing, 

 applicable to the Gold Bullion Assay. 



By George Foord, Esq. 



[Read 8th November, 1875.] 



It is well known that the modern practice of the gold 

 assay includes many refinements upon the routine formerly 

 practised ; that it commands chemicals of greater purity, 

 balances of greater sensitiveness and accuracy, and certain 

 accessory tools which facilitate and expedite the work. 

 But it is also well understood, especially by those imme- 

 diately interested in the work, that although the method has 

 been thus refined, it still falls short of absolute certainty and 

 exactitude, and that further progressive improvement is still 

 possible. One of the chief sources of minor errors which may 

 yet be in part removed is that which belongs to the opera- 

 tion of weighing, and it is the object of this paper to propose 

 modifications, Avhich promise some advantage over the 

 method of weighing hitherto practised. 



It will assist my explanation if in the first place I offer a 

 few practical observations concerning the construction and 

 adjustment of the assay beam. 



For weighing, in the ordinary routine of the gold assay, 

 there are required : — 



1st. — A sensitive and accurate balance. 



2nd. — A series of accurate weights. 

 The balance-beam should be as light as possible, the requisite 

 stiffness being at the same time secured. The arms of the 

 balance should be of exactly equal length, measuring from 

 the centre to each end knife edge. The knife edges must be 

 sharp. The centre of gravity should be a little below the 

 plane of the three knife-edges, and immediately under the 

 centre knife-edge. Each arm of the beam should be accu- 

 rately and distinctly divided by lines into fifty parts, each 

 fifth mark from the centre, however, being distinguished by 

 a dot instead of a line ; these dots will therefore indicate 

 tenths of the space between the centre and end knife-edge. 

 The pans, of whatever pattern, should be small and light. 

 The scale, measuring the sweep of the index or needle, 

 should be accurately divided, so that each division of this 

 scale represents uniformly a subdivision of the smallest 



