ON GOLD-ALUMINIUM ALLOYS. 217 



ordinate crystallisation of the same body, the original outline will become lost, and 

 the final outline will be rounded as we see it in the photographs. Prolonged etching 

 of the 18'1 atom alloy breaks up the surface of the blobs and partially reveals a 

 structure that may l>e that of an original skeleton, although this structure is not 

 altogether what we should expect in such a case. 



We will now consider the alloys taken in order, starting from the pure gold end of 

 the curve. 



The first alloy examined carefully was that containing 1 '27 per cent, by weight of 

 aluminium, that is, 8 '56 atoms. It has to the eye the appearance of gold, is soft, and 

 does not polish well, and the unetched surface shows no detail. Etching with bromine 

 water produces a very brilliant surface, and a 2 millims. immersion objective, with a 

 power of 500 diameters, now shows it to be made up of approximately hexagonal 

 polygons, apparently of gold, with an incomplete network of fine brown lines between 

 them. We do not give a figure of this as it is identical with types that occur later in 

 the curve, for example at E and H. BEHRENS, ARNOLD, ANDREWS, and all who have 

 studied dilute solutions of one metal in another, have observed the same type of 

 structure. 



The alloy with 12 '5 atoms similarly treated shows the polygons of gold somewhat 

 rounded at the angles, and surrounded by much more of the brown mother-substance. 

 This brown substance has a minute sparkle in it when examined with a power of 500. 

 A power of 1500 and careful focussing showed this sparkle to be due to numerous 

 spots, which sometimes appeared white ; this is presumably the detail of the eutectic. 

 In the photograph (fig. 1 ) the darker parts are the spaces full of eutectic, but the scale 

 is too small to show the smaller detail. 



The next alloy contained 16'9 atoms of aluminium. It was very white when 

 polished by the wet method, and the microscope showed, before etching, golden crosses 

 dimly visible on a white ground. When lightly etched, a power of 50 diameters 

 brought out a beautifully regular pattern of rectangular crosses and bars on a dark 

 ground that was finely mottled with gold. The ground was brown and the pattern 

 generally golden, but under some conditions of etching the pattern seemed to be 

 white, and many of the minute specks in the brown mother-substance seemed to be 

 white also. A power of 1000 brought out very clearly the fact that the ground was a 

 eutectic mixture. A slowly-cooled alloy, containing 18'1 atoms, of which we give a 

 photograph (fig. 2), presented a very similar pattern, but on so large a scale that it was 

 visible to the naked eye ; the pattern in the slowly-cooled alloy was, however, always 

 golden. Although we have mentioned the occasional whiteness of the pattern in the 

 16 "9 atom alloy, we do not feel able to attach any importance to it, as it was probably 

 due to some electrolytic action occurring during the process of etching. 



Light etching with bromine water shows very clearly that the ground of the 

 18'1 atom slowly -cooled alloy is a eutectic mixture, as the gold is left standing up in 

 bent rods and crinkles while the other substance of the eutectic is eaten away. We 

 VOL. cxciv. A. 2 F 



