13^ J. Scully — On the Efech produced hy small quantities of [Ko. 2, 



owing to brittleness, if they contain only 3 per mille of bismuth ; while if 

 the latter metal forms less than 0*5 per mille of the whole miass the 

 ductility of the bars is not much affected. 



In conclusion, I have much pleasure in recording my appreciation 

 of the services of Messrs. J. R. L. Durham and E. Hood, Head and Se- 

 cond Assistants in the Assay Office, in carrying out under my directions 

 many details of the experiments recorded in this paper. 



Calcutta, March y 1887. 



P. S. — June 10th. The experiments detailed in the foregoing paper 

 were completed early in l!^ovember 1886 and it was proposed to embody 

 the conclusions formed in an official report to be submitted this year. 

 It was suggested to me, however, that the subject investigated might be 

 of general interest, and I had determined to publish this paper when 

 I noticed in the Chemical News of March 21st, 1887, (p. 137), a short ab- 

 stract of a paper on " Silver containing Bismuth " by Messrs. Gowland 

 and Koga of the Japan Mint. I have delayed presenting this paper 

 until I had read the full text of the communication from Japan (Journal 

 of the Chemical Society No. CCXCIV, May, 1887, p. 410), and I may 

 now make a few remarks upon it in connection with what I have 

 advanced. 



Messrs. Gowland and Koga's very interesting paper to a great extent 

 covers ground which I had not investigated, viz., the want of uniformity 

 in composition of silver bullion containing bismuth. This part of the sub- 

 ject was suggested to me for experiment, as some of my results seemed 

 to show that bismuth mixed with silver by melting and careful stirring 

 does not diffuse itself evenly throughout the solidified mass. But that 

 fact, now proved by Messrs. Gowland and Koga, was of no practical impor- 

 tance to us in the assay aud valuation of bullion, seeing that it is an 

 invariable rule now in the Calcutta Mint to premelt and assay by a 

 granulated sample every kind of bullion tendered to the Mint — from refined 

 bars 999*5 fine to Mexican Dollars. The well-known want of homogeneity 

 in solidified silver-copper alloys, and other contingencies to which 

 silver bullion is subject, render this course imperative for purposes of 

 valuation on any extended scale. The cutting of samples from silver 

 bullion for assay, even if the spot where such samples should be cut has 

 been determined after most laborious investigation, can at best give 

 merely approximate results ; as indeed our authors admit for the case of 

 silver containing bismuth. 



With regard to the toughening or refining of silver containing 

 bismuth, Messrs. Gowland and Koga mention that this operation is 

 successfully performed in the Japan Mint by prolonged exposure of the 



