applicable to the Gold Bullion Assay. 109 



One point may be added respecting the wear of weights, 

 in special reference to those of platinum, namely, that when 

 worn they may be easily restored to accuracy, by taking a 

 small piece of fine gold enough to make the particular 

 weight slightly heavy ; if the gold be placed on the weight 

 and heated before the blowpipe the gold on fusing will 

 attach itself firmly to the surface of the platinum, and the 

 weight may now be tried and reduced alternately, reducing it 

 first if necessary on a very fine file, and finally rubbing it on 

 a hone or touch-stone until the precise weight is arrived at. 



Of course there should be a standard weight, and copies 

 of it with which the ten-grain weight can from time to time 

 be compared. A ten-grain standard weight of rock-crystal 

 offers the advantages of great hardness and inalterability ; 

 but as the displacement of air by equal weights of bodies 

 differing so much in specific gravity as quartz and platinum 

 is considerable, and may cause under extreme variations of 

 atmospheric temperature, pressure, &c, a difference of 

 apparent weight amounting to nearly two-thirds of a ten- 

 thousandth of unity, equal to sixty-six pounds in a million 

 sterling, this influence must either be allowed for by 

 correction in such comparisons of quartz and platinum 

 weights, or the comparisons must be always made under the 

 same atmospheric conditions. For the preparatory weighing 

 of the samples for assay, conducted usually by an assistant, 

 and for which the wear of the weight is likely to be com- 

 paratively great, a quartz weight will be found peculiarly 

 suitable. This quartz pound is counterpoised by a weight 

 in the opposite pan, adjusting the two by the tongue of the 

 beam until equilibrium is obtained ; the quartz weight is 

 then removed, and the gold samples to be weighed substi- 

 stuted, one after the other, for it ; this method gives the 

 advantages of double weighing, and if the beam be suffi- 

 ciently sensitive, very uniform weighments will be effected, 

 even though the arms of the balance be of irregular length. 



, , DESCRIPTION OF THE PLATES. 



Plate 1 snows : — 



1 . The series of platinum weights in common use by bullion assayers, 



ranging from -80 to -99 inclusive; arranged in partitioned tray. 



2. Proposed series of "compensating" weights, ranging from -80 



to -98, the rider representing -99 ; arranged as the foregoing. 



Plate 2 : — Shows the assay beam, with the notation of its major divisions 

 according to the compensating method. The additional end piece, as 

 described in the paper, is shown at (a), and a side view of^t-at (b), in which 

 its inclination from the plane of the beam itself (c) is represented. 



The index reel (d) is shown full size at (e). 



Por clearness the supporting columns of the balance and other accessories 

 are omitted in the drawing. 



