1903.] The Properties of the Aluminium-Tin Alloys. 279 



which we examined makes the freezing-point curve connecting these a 

 peculiar one. With the exception of a slight deviation which is 

 apparent at the point B (see curve Diagram 1), it slopes uniformly 

 downwards until about 25 per cent, of aluminium is present (point C 

 on curve). From this point, however, until the percentage of 

 aluminium reaches 17 per cent. (D on curve), the temperature at which 

 the first break occurs remains practically constant. Indeed, when 

 uniformity in the conditions of cooling is maintained, the alloys from 

 25 per cent, aluminium to 17 per cent, aluminium may be said to 

 begin the separation of solid matter at almost exactly the same 

 temperature, namely 561 C. When the percentages of aluminium 

 are less than 17, the downward trend of the curve is resumed at a 

 more rapid rate than before, with perhaps a slight further deflection at 

 or about the point E in our diagram. 



In our experiments the flattening of the curve extended from the 

 alloy having the composition 24'8 per cent, aluminium to that con- 

 taining 17-43 per cent. Between these limits the first signs of heat 

 evolution appeared at from 563 to 560 0> 5 C. A definite compound of 

 the formula AlSn would have the composition 



Al 18-5 per cent. 



Sn 81-5 



Whether the solid matter separating out at the temperature stated 

 has this composition or not, it is at least the case that in that part of 

 the curve to the left of this point the separation is taking place from a 

 solution containing more than one atom of aluminium to each atom of 

 tin ; to the right it is taking place from a solution in which atoms of 

 tin preponderate. The homogeneous liquid mixtures, therefore, to the 

 left are solutions of tin in aluminium, those to the right are solutions of 

 aluminium in tin. 



A freezing-point curve for these alloys has already been given 

 by H. Gautier.* His results differ from ours most markedly in 

 this, that in the region CD he finds the separation of solid matter 

 to commence at temperatures higher than is the case in alloys 

 both to the left and right of it. The result is a very pronounced 

 " kink " in the curve, which in our experiments is represented 

 to the extent of only 1 C. in the case of the alloy containing 

 17-43 per cent, of aluminium. The explanation of the difference is 

 doubtless to be found in the different rates of cooling in the two series 

 of observations. For the sake of comparison we append the figures. 

 The melting point of aluminium is given by Gautier as 650 C. ; we 

 have chosen the figure 655 C. 



' Comptes Kendus,' vol. 123, 1896, p. 109. 



