368 The Gold-Aluminium Series of Alloys. [Jan. 28, 



at a point which has yet to be definitely fixed, but which is several 

 degrees above gold.* In fact, when workmen who are accustomed to 

 melt gold on a large scale, attempt to melt this purple substance, they 

 find it difficult to believe that they are dealing with a gold alloy, as 

 it is so infusible. 



The melting points of the rest of the series richer in aluminium 

 appear to fall continuously to 660, a little below the melting point of 

 aluminium (665 C.). 



The purple alloy presents the only case, known to the author, of an 

 alloy, free from mercury, having a higher melting point than that of 

 the least fusible of its constituents, and he considers that this fact 

 affords strong evidence of its being a true compound of gold and 

 aluminium. 



It is generally admitted that there are true compounds in the 

 copper- tin series, for SnCu 3 and SnCu 4 seem to be well defined, but 

 their melting points are much lower than that of copper. 



A. P. Laurie has just shown ('Phil. Mag.,' January, 1892) that in 

 the gold-tin series, the alloy containing 63 per cent, of gold and 

 37 per cent, of tin has an electromotive force which distinguishes it 

 from the rest of the series and points conclusively to its being a true 

 compound, but the author finds that it melts readily below redness. 



The melting points of ordinary chemical compounds are often 

 much higher than the melting point of the least fusible constituent. 

 Galena, for instance, melts at a strong red heat ; it is difficult to fix 

 the point accurately as the substance volatilizes, but it is close to 

 900 C. Its constituents, lead and sulphur, melt at 335 and 115 

 respectively. Stibnite also, sulphide of antimony, melts at about 

 530, according to Dr. Joly, while antimony fuses at 440. 



The gold-aluminium series is of unusual interest, and well deserves 

 careful attention. 



* [Two very careful experiments were made, each with 40 grammes of the 

 alloy, the cooling curve being traced by the autographic recorder already 

 described (' Roy. Soc. Proc.,' loc. cit.}. These curves gave 1065 and 1070 respec- 

 tively as the melting point of the alloy AuAL, the mean of which is 32'5 higher 

 than the melting point of gold. If, however, small quantities of the alloy be fused 

 before the oxy-hydrogen blowpipe, it is easy to obtain a lower result, as aluminium 

 is readily burnt out from the little mass. The composition of the alloy is thereby 

 changed to one of the series richer in gold, of which the melting points are lower 

 than that of gold. Feb. 9, 1892.J 



