MESSRS. C. T. HEYCOCK AND E. H. NEVILLE ON 
an hour. But in the case of Sn 27 a slow heating of several hours was tried ; it gave 
exactly the same pull-out temperature as a much more rapid heating’ had given. 
The results for pure metals are stated in the third column of the table. They are • 
in veiy fair agreement with the freezing-points. It is rather curious that the result 
for aluminium should be practically the same as the freezing-point. We then applied 
both methods to a few alloys. The results are shown below, the breaks by short 
horizontal lines, and tlie ])ull-outs l)y round dots. It is evident that the allovs 
remain practically rigid up to temperatures well above our solidus, indeed nearly u]) 
to the hcpudus. We infer frc)m these results that the alloy gives way, not when the 
eutectic, or otlier mother-substance, between the primary combs liegins to melt, l)ut 
at considerably higher temperatures, vdien a good deal of the primary has itself been 
dissolved. The skeleton crystals of the solid phase must form a sufficiently rigid 
framework to prevent the breaking of the bar, or even the pull-out, although they are 
surrounded by liquid. From the nature of the e(pulibriuni in the region between the 
solidus and the liquidus it is quite conceivable that the rigid scaffolding might become 
more infusible by the draining away of the liquid fi’om the bars. Thus the 
temperature we were measuring was necessarily indefinite, hut depended more on the 
melting-point of the solid phase than on the commencement of liquefaction. 
ihe experiments failed to throw any light on- the })Osition of the solidus, hut they 
have proved that the skeletons of the solid phase are, even at the tempeiatures 
between tlie solidus and the hlpiidus, rigid structures and not plastic masses. This 
fact, together with the photographs which indicate when the solid phase fills the 
ingot, entitles us to say that below the solidus Khlcmdef, as drawn by us, the alloys, 
though they may be at a bright red heat, are rigid solids. Hence any crystallisations 
taking ace below the solidus occur in an already solid and crystcdline body. 
