Pyrometer on the Melting -points of Gold end Silver. 227 



method at the Mint, in conjunction with Professor Roberts- 

 Austen, on the effect of gold dissolved in silver. We found 

 that the addition of 0*6 per cent, of gold raised the freezing- 

 point of silver by about o, 2 C. A further addition of 

 0*6 per cent, raised the freezing-point o, 6 C. in all. A 

 further addition of 1*2 per cent., making 2*4 per cent, in all, 

 raised the freezing-point l c, 3 0. It is interesting to observe 

 that the effect of gold on silver is similar to that of silver on 

 gold, and forms an exception to the general rule. It might 

 have been expected, a priori, that the addition of 1*3 atoms of 

 gold per 100 of silver would have lowered the freezing-point 

 by about 17° C; instead of which, a small but clearly marked 

 rise was observed. Silver, as shown above, when dissolved in 

 gold, produces a similarly slight but distinctly appreciable 

 fall. It may be remarked that the rise produced by gold in 

 silver bears nearly the same proportion to the interval 

 between their freezing-points ( ( J45° and 1037°) as the num- 

 ber of atoms of gold present bears to the number of atoms of 

 silver. 



As a further experiment, aluminium was added to the mass 

 in proportion (about 0*6 per cent.) sufficient to form, with 

 the gold already in the silver, the alloy AuAl 2 , which Roberts - 

 Austen has shown to possess most interesting and remarkable 

 properties. The mass was well stirred, and a curve of cooling 

 was taken, shown by the dotted line (No. 6) in Plate V. The 

 temperature fell to 819° Pt., where it remained quite stationary 

 for about a minute. It then fell slowly for about two 

 minutes, before the normal rate of cooling was reached. 

 The lowering of the freezing-point 12*3° G. is rather less than 

 that required by theory for a solution of 1*3 molecules of 

 AuAl 2 in 100 of silver ; but it was observed that an appre- 

 ciable quantity of the alloy sublimed about the mouth of the 

 furnace and on the stem of the pyrometer. It is also doubtful 

 whether the mass was perfectly homogeneous. The extent to 

 which the freezing-point was lowered would, however, appear 

 to indicate that this was a true case of solution. 



From want of sufficient metallurgical knowledge, I cannot 

 pretend to interpret these results more fully. I am induced 

 to publish them, incomplete as they are, rather as an illustra- 

 tion of the degree of accuracy attainable by the use of platinum 

 pyrometers, and in the hope that, in more experienced hands, 

 they may lead to really valuable results. 



I have also tested one of these pyrometers in molten east 

 iron raised to a temperature considerably above its melting- 

 point. As the mass cooled it assumed a pasty condition, and 



