On the Results of Chilling Copper- Tin Alloys. 



173 



make the alloys in an atmosphere of coal-gas or hydrogen, and to 

 allow them to cool in this atmosphere. If made in this way, we found 

 that all alloys, from A almost to D, showed on the top of the ingot a 

 regular crystallisation in relief, of the rectangular comb-like character 

 so often seen on the surface of cast metal. This was as perfect in the 

 white metals between C and D as in the red alloys between A and B. 

 These crystals disappear when the point D is reached, although with 

 much more tin other types of raised crystals are seen. These combs 

 are of course primary crystals, standing out on account of the con- 

 traction of the solidifying mass and the consequent retirement of the 

 mother liquid. When the ingots of alloy are cut, the surfaces polished, 

 and the internal pattern brought out by ignition or etching, one sees, 

 as Charpy and Stead have shown, that similar combs, rich in copper, 

 occur in the interior of the ABC alloys, the combs being embedded in 

 a matrix which is itself complex (see photo. 1, PI. 2). These combs are 

 numerous and large in the gun-metals of the region AB, but decrease 

 in numbers, size, and perfection as we approach C. For some distance 

 to the left of C they are much broken and distorted, and to the right 

 of C they do not appear at all in the body of the alloys ; but they 

 exist on the outside in the same perfection as before. Moreover, if 

 the top of one of the alloys anywhere between a point a little to the 

 left of and the point D be slightly ground down so as to obtain 

 sections half through the raised crystals, and the pattern examined, it 

 is found that the crystals are not homogeneous, as one would expect a 

 crystal to be, but that each crystal is full of a well-marked pattern 

 identical with that of the body of the alloy. To illustrate this pecu- 

 liarity, we give a photograph of the top of the alloy containing 14 

 atomic per cents, of tin (photo. 2). Hence it appeared that the alloys 

 underwent remarkable changes both during and after solidification. 

 In the alloy of photograph (2) the larger detail in the substance of the 

 bars of raised crystal, or something not unlike it, was formed before 

 the raised pattern, but the smaller detail, hardly seen at this magnifi- 

 cation, is more recent than the raised pattern. 



Photograph (1) shows the large primary combs existing in the 

 interior of an alloy containing 12 atomic per cents, of tin, and photo- 

 graph (3) shows the utterly different pattern existing on the other 

 side of C. It is that of an alloy containing 16 - 7 atomic per cents, of 

 tin. It must be remembered that on the outside the alloy still shows 

 the combs. These alloys were slowly cooled, that is, not subjected to 

 any sudden chill during cooling. A pattern like that of photograph 

 (3) is given by Gharpy for an alloy containing equal weights of copper 

 and zinc. We have also found it in some silver-zinc alloys, and we 

 think it always means that changes have taken place in the solid 

 alloy. 



The patterns at all points on the curve were so puzzling that we 



2 



