134 



REPORT — 1900. 



a simple formula — for example, in the AgCu and AuCu curves. Sir George 

 Stokes has lately recalled our attention to this point. Paterno and 

 Ampolla (•^) have noticed a number of similar cases in organic 

 mixtures. 



A good many curves indicating by intermediate branches the 

 existence of compounds have now been determined, but for the purpose 

 of illustrating the subject further the curve of AuAl will be taken. 



As can be seen from fig. 2, there are seven branches, each corre- 



FiG. 2. 



Temp 



Cent. 

 IIOO 



1000 



8001 



800 



700 



600 



fiOO 



Weight per cent, of Alumiuiuui 



50 



100 



sponding to the crystallisation of a different solid. The extreme branches, 

 AB and IJ, being regarded as those of the two metals, we have five left, 

 each of which Tnay indicate a compound. The branch def has its 

 summit exactly at the formula AuoAl. The microscope shows that the 

 summit alloy e is an almost homogeneous body, and that all solid alloys 

 whose composition lies between that of d and f contain large crystals of 

 the E body immersed in a mother substance. As we descend the curve 

 from the summit the large crystals of E are found to occupy less and less 

 of the whole alloy, until at d and f they cease to exist. Exactly similar 



