OiST THE STRUCTURE OF METALS. 
4 27 
in all our specimens, except in the alloy with potassium. # When revealed by the 
action of aqua regia, these joints usually appear under the microscope as being of a 
brick-red or deep purple colour, and are of variable thickness. In certain cases they 
are only mathematical lines, that is to say, without thickness, and their existence is 
only revealed by the different coloration of the grains of which they form the limit (see 
photos. 12, Eh, 150 diameters ; and 9, Eh, 1000 diameters). In other cases, the 
joints revealed by etching have a definite thickness (see photos. 15, T1; 14, In ; 11, K ; 
150 diameters). These joints, which are more or less thick, when carefully examined 
under high magnification, may be represented by two types. One of these is formed 
of a sunk line, generally very fine, on the edges of which crystals, apparently little 
cubes, are ranged, which are secondary crystallizations. These are sometimes isolated, 
but more usually run together, and are joined in continuous files (see photo. 17, In, 
500 diameters). Other joints are simply constituted by the separation for a short 
distance of primary crystals, between which the acid can penetrate more easily and 
etch away a deeper channel. We will call the latter, as distinguished from the 
former, non-crystalline. t 
Considered from the point of view of the nature and the thickness of their joints,| 
our thirteen specimens may be divided into five groups. 
1st. Pure gold; alloys with Zr, Al, Eh, Pd, Zn ; joints which are usually non- 
crystalline, of a thickness which equals 0 to 2 /x (p, = O'OOl mm.), the mathematical 
joints or very fine ones are dominant. 
2nd. Alloy with lithium; the thickness of the joints equals 0 to 3 ‘5 /a ; the thick 
joints are crystalline but are rare and disconnected, the mathematical joints, or very 
fine joints are dominant. 
3rd. Alloys with Sb, Se ; crystalline joints which are often broken ; the thickness of 
these equals 1 /x to 3'5 /x, or a mean thickness of 2 /x. 
4th. The alloys with In, K ; crystalline continuous joints, with a thickness of 1 to 
4 /x, or a mean thickness of 2 to 2‘5 /x. 
5th. Alloys with Tl, Bi; crystalline joints, which are continuous, often passing into 
* The alloy of gold and indium lias not been etched by sulphuric acid. 
t It may be well to record an ajipearance which is sometimes met with, and may lead to error. 
Certain thick non-crystalline joints simulate the presence of an interposed cement, for, on account of 
the difference of level between the depth of the joint and the surface, shadows are cast. This is most 
clearly shown by the photograph 10 (1000 diameters). The plane etched by aqua regia had not been 
repolished after the previous etching by sulphuric acid. It looks as if a film of foreign substance 
was interposed between two neighbouring grains, but, on looking at it closer, it is evident that the 
traces of this film join those of one of the grains, and it is not possible to admit a doubt that we have 
merely to deal with the structure of pure gold. 
+ By the expi’ession “ thickness of a joint,” we understand, in the case of the crystalline joints, the 
sum of the joint properly so-called, and of the crystalline border, which it is usually impossible to 
separate from it. We have, in fact, measured in all cases the thickness of the dark line. It must not 
be forgotten that it is a question of the apparent dimensions, which are much exaggerated by the 
etching. 
3 i 2 
