470 Mr. E. Matthey. [Jan. 20, 



e iven off freely from the arsenical bismuth until a temperature of 

 *95 is reached So that the temperature of disassociation of 

 alloy, containing 0'65 per cent, of arsenic, is 114 C. higher than the 

 nelthiff point of the mass. It was interesting to determine at what 

 temperature the arsenic would be evolved if the alloy were heated in 



'^portion of the bismuth alloy containing O65 per cent, of arsenic 

 was introduced into a hard glass tube, slightly depressed to its centre, 

 and connected at one end with a Sprengel pump, by which it could be 

 rendered vacuous. 



The pyrometric wires were in contact with the metalli 

 passed to the galvanometer through the opposite sealed end of the 



glass tube. 



Heat being applied, the first indication of the volatilisation of the 

 arsenic, shown by the condensation of a film on the cool part of the 

 tube, occurred at 275 C. When the alloy was quite melted the tern- 

 perature indicated was 316 C. ; arsenic came off freely when the 

 temperature rose to 569 C., condensing in a black mirror. 



The metal was then allowed to cool, and its setting point was 

 found to be 268 C., which corresponds with that of the melting point 

 of bismuth. 



The bulk of the arsenic does not appear to be evolved in vacuo at a 

 lower temperature than in air. 



As regards the industrial application of the process, some ten to 

 twelve tons of arsenical bismuth have already been treated in this 

 very simple way, and it has been satisfactorily ascertained that there 

 is no loss of bismuth by volatilisation with the arsenic. 



PART V. The Separation of Bismuth from Antimony. 



The process hitherto adopted in practice for the separation of 

 antimony from bismuth has usually consisted in a simple fusion at a 

 dull-red heat with bismuth oxide or bismuth " litharge " an opera- 

 tion successful enough as to its results, but one requiring no small 

 amount of skill in manipulation; it is also one by which only small 

 quantities can be treated readily at one time and moreover, the 

 temperature which is necessary to effect the separation of the 

 antimony involves appreciable loss on account of the volatilisation of 

 the bismuth at a red heat, notwithstanding many tens of tons of 

 bismuth have, however, been treated under my direction by this 

 process. 



In an operation lately conducted, involving the melting of a 

 quantity of bismuth containing about one per cent, of antimony, it was 

 found that a peculiar oily film was noticeable rising to the surface of 

 the melted alloy; this film did not form all over the s-irface of the 



