ON GOLD-ALUMINIUM ALLOYS. 219 



shows also another feature of both sections. This feature, of which we cannot at 

 present see the explanation, consists in a much coarser polygonal structure, the whole 



a iv; i Ix-iii^ di\ iilcil l.y ilark l.amU ini ..-,],, ,,,-. ,-ai-li .>t'uliir|i OOntailtf t'l" 111 t.-n I., lil'iy 



of the smaller gold-edged cells. The width of the dark bands is about that of an 

 average sized ^old-fd^ril cell, ami the material of the bands seems to be that of the 

 inside of the cells. Photographs of the 19 '8 alloy, taken with oblique illumination, 

 show the larger polygonal structure, but only very imperfectly that of the gold-edged 

 cells ; we, therefore, give a drawing of this alloy (fig. 5). These sections, when 

 magnified 200 diameters, very much resemble the figures given by Professor ARNOLD 

 (' Engineering,' February 7, 1896) for gold containing 0'2 per cent, of bismuth or 

 silicon, that is, for a nearly pure substance. 



Whatever the meaning of the larger polygonal structure may be, our immediate 

 point is that the cells of dark material, for the most part isolated from one another, 

 are primary crystals, probably of a body pure near B. Both of these sections are 

 far more intelligible under the microscope than in the photographs. With careful 

 focussing and a power of 1000 one sees that the golden network of the 19 '8 alloy is, 

 where it widens at the angles, a eutectic mixture of gold and dark not uidike the 

 drawing, only far more minute. 



The 19 '8 atom alloy of flg. 5 was not one of the extracts made during the determina- 

 tion of the freezing point-curve, but was specially prepared for microscopic examination, 

 it was, however, extracted from the crucible by sucking up in the usual way. We have 

 since polished a section of the ingot left in the crucible after extracting this alloy. 

 The ingot, of course, cooled much more slowly than the rod extracted. The polished 

 section of the ingot shows before etching a cellular pattern, the cells being full of a 

 uniform material, while the intercellular matter, or mother-substance, of which there 

 is a good deal, broadening at the angles, is plainly a fine eutectic (fig. 6). A power 

 of 400 diameters shows this well. This section, even when lightly etched with 

 bromine, shows no trace of the larger polygonal structure. Both these alloys polished 

 to very white surfaces. The rod of alloy from which the latter section was cut was 

 white and brittle with a conchoidal fracture. 



At 2T1 atoms we have, after bromine etching, the same brown polygons forming the 

 mass of the alloy, but now the interstices are filled with a new white body, with 

 which we shall become familiar later on. We give a drawing, under oblique illumina- 

 tion, of this alloy (fig. 7). The drawing accurately represents a portion of the surface, 

 but is rather misleading as to the amount of interpolygonal matter, for if one takes a 

 general survey of the surface one sees that there is very little of this. We give a photo- 

 graph (fig. 8) to make this fact clear. Hence the alloy is extremely near to being a pure 

 substance, nearer, for example, than the previous alloy. After the drawing was made, 

 the section was subjected to a prolonged etching with strong bromine water. This 

 brought out very clearly the white interpolygonal substance standing in relief above 

 the level of the brown. This alloy before polishing was brassy-white with a 



2 F 2 



