On the Results of Chilling Copper-Tin Alloys. 



177 



the solidification of the whole mass to mixed crystals, which, assu- 

 ming no lag in the transformations, should be uniform. The long 

 slope OP would then correspond to the cooling of a solid mass of 

 uniform crystals, and therefore the alloys chilled in this region of 

 temperature show no pattern. But at P the solid solution becomes 

 saturated, and on cooling below this point the band and fern leaf crys- 

 tallises out. At a still lower temperature, probably Q, the mother 

 substance of the fern leaf breaks up into a eutectic, formed in the 

 solid. We think that P is a point on Austen and Stansfield's lower 

 curve, and that Q is the eutectic angle of that curve. It will probably 

 be found that the mother substance in all alloys from about B to D 

 breaks up into a complex when the alloys cool to the temperature Q, 

 so that if cooled slowly it is a eutectic, but if chilled above Q a 

 homogeneous body. 



It is not difficult to form a conception of how the type of pattern 

 found below the temperature P originates. Slightly above the tem- 

 perature the alloy consisted of crystal grains surrounded by mother 

 liquid somewhat richer in tin. At the moment of complete solidifica- 

 tion the grains should have adjusted themselves so as to be identical 

 throughout, but it is improbable that so perfect an equilibrium was 

 attained, and the solid mass at temperatures below must have con- 

 tained nuclei richer in copper than the material surrounding them. 

 In fact, prolonged polishing brings out a vague pattern in relief, 

 showing differences of hardness, and therefore of composition. Now 

 the alloy that we are considering lies to the right of Austen and 

 Stansfield's eutectic angle in their lower curve ; hence when the solid 

 solution became saturated the new crystallisation commenced in the 

 interspaces rich in tin, and more or less took their form. It is clear 

 that the resulting structure would in section give the bands and poly- 

 gons of the slow-cooled alloys. Similarly the inclusions of mother 

 substance in the grains existing at would be the origin of the isolated 

 fern leaf. 



Although it was hardly necessary, we thought it would be interest- 

 ing to arrive at the condition of no pattern, starting from the solid 

 alloy instead of from the liquid. We therefore took a fragment from 

 an ingot of the same slowly cooled alloy, heated it to a faint red heat 

 in the Bunsen flame, and dropped it into water. It showed no pattern 

 after being polished and ignited to a pale orange. It was then heated 

 to a temperature a little below redness, and allowed to cool for five 

 minutes above the flame, repolished, and brought to the orange state. 

 It then showed a very perfect slow-cooled pattern, the fern leaf being 

 particularly good. The polygons appeared to be of the same size as in 

 the original alloy, which had taken an hour or more to cool, but the 

 bands were much thinner and the fern leaf smaller ; the eutectic also 

 was very scanty, while in the original ingot there were large spaces of 



