On the Results of Chilling Copper- Tin Alloys. 



175 



of from 5 to 10 grammes of the alloy, contained in little test-tubes of 

 Jena glass, were immersed in the bath ■ these were in an atmosphere 

 of coal-gas, and so did not oxidise. The bath of tin was then allowed 

 to cool slowly and uniformly, and when the temperature fell to one of 

 the selected points, a tube was taken out and plunged into water. 

 The alloy was thus chilled, the slow cooling being brought to an 

 abrupt end at any desired temperature. 



The chilled alloys were afterwards ground down and polished in the 

 usual way. After the trial of many reagents for bringing out pattern, 

 we adopted the method of slightly heating the surface until the film 

 of oxide formed was of a pale yellow colour. Behrens some years 

 ago recommended this method, and Mr. Stead has pointed out that 

 it develops differences of chemical composition very well, while 

 etching reagents complicate the picture by revealing the orientation 

 of crystals and other details which are not always needed. With 

 one or two doubtful exceptions, we find that in alloys richer in 

 copper than Cu 3 Sn, the parts which oxidise most rapidly, and are 

 therefore darkest in the yellow stage, are the softer parts contain- 

 ing most copper. When alloys on the branch ABC are oxidised the 

 pattern is very distinct to the eye, but it is sometimes difficult to 

 obtain much contrast in the photographs ; in such cases (for example, 

 in the alloy of photograph 1) we etched the surface with strong 

 ammonia, which also darkens the parts richest in copper. Alloys on 

 the branch ABC are very sensitive to reagents such as ammonia or 

 hydrochloric acid, and from C to D, where these have but little 

 action, a mixture of hydrochloric acid and potassium chlorate etches 

 rapidly. One can use these reagents to control the effect of heat 

 oxidation in cases where the low temperature of chilling makes it 

 possible that the heating needed to produce the yellow colour may 

 have reversed the result of chilling; but we find that there is not 

 much danger of such a reversal. 



The upper point alloy, chilled at the commencement of solidification, 

 was generally found to be granulated by the operation of dropping 

 into water, but portions could always be found suitable for polish- 

 ing; the other alloys had always solidified before the chilling, and 

 therefore gave compact ingots. 



After polishing, the alloys were heated until a pale yellow oxidation 

 colour was produced on the surface. 



Alloy (1), chilled when much of the metal was still liquid, shows a 

 pattern of large primary skeletons, more or less comb-like in appear- 

 ance, which oxidise much more rapidly than the mother substance, 

 and which therefore contain more copper than it (photo. 4). 



Alloy (2), chilled when the solidification was almost complete, shows 

 skeletons much softer in outline and not differing much in oxidation 

 colour from the ground ; but these skeletons occupy a much larger 



