66 
MESSRS. C. T. HEYCOCK AND F. H. NEVILLE ON 
probaljle that they are due to a contraction accompanying the reaction in whicli the 
H is formed from the mixture of p and liquid. 
The chill of Sn 42 (fig. 96), and in a still more marked degree the chill of Sn 40 
(fig. 97), contains p uniformly scattered through the massive H. This study of the 
H transformation seems to point to the conclusion that the H body has at 350° a 
composition not far from that of Sn 45. This conclusion is confirmed by the results 
of analyses of the H extracted from ingots of Sn 90, which give the value of Sn 45‘5. 
Thus neither the microscope nor the analyses allow us to regard H as pure CuSn, but 
it is very probably this body with a few per cent of CugSn in solid solution. 
Fig. 98 is a 2 )hotogTaph of an ingot of Sn 60 after 21 days in mercury at 350°. 
Like most of the ingots boiled in mercury it was not chilled, although it was cooled 
rapidly at the end of the period of boiling. When compared with Sn 50 we see less 
of the H and more of the eutectic. The progress of the masses of H towards the 
condition of polyhedra is ^vell marked. 
Fig. 99 is an ingot of Sn 85 after 21 days in mercury. Some of the masses of H 
have the rounded form that indicates their origin from t], but there are other crystals 
of H whicli are probably idiomorphic. 
Fig. 101 illustrates tlie facility with which the reverse change 
H = p -f liquid 
takes place. It is an ingot of Sn 50 that had been heated for 10 days in mercury 
vapour, the p having been thus almost entirely removed; it was then heated for a 
few minutes to a temperature a little above 400°, with the result that all the masses 
of H have broken up into p and liquid. The pattern, resembling that of 93, produced 
by the mercury boiling, can still, liowever, be traced. 
Fig. too records an early attempt to study the reaction ol the H temperature. A 
considerable mass of the alloy Sn 90 was maintained for several hours at a temperature 
slightly below 400°, and it was automatically stirred by means of a massive copper 
stirrer made of coils of thick wire. The reaction, which produced nothing but H, was 
continued until the stirring was brought to an end by the large quantity of solid H 
that liad formed. The ingot was tlien cut and polished as usual. A comparison with 
fig. 89 makes the great increase in the amount of H veiy evident. 
The Isolation and Analysis of the Bodies y and H.—The fact observed during the 
etcliing of the alloys, tliat hydrocldoric acid has no perceptible action on the crystals 
of Tj or H, can be utilised for tlie separation of these bodies from an alloy in which 
one of them is present together with tin. * If we wish to get pure p, the alloy must 
l)e sharply cliilled at a temperature above 400°; to obtain pure H, the alloy must 
have 1jeen maintained at a temperature a little below 400° for a period long enough 
to transform all the y. It is in tlie observance of these two precautions that our 
method f)f isolating the substances constitutes an advance on previous work. We 
find that the ingots of alloy are rapidly disintegrated by the action of strong 
