551 
of Edinburgh , Session 1881-82. 
the molten metal. A third view is that the molten metal on 
solidifying expands , and that it thereafter contracts during cooling 
until it reaches the ordinary atmospheric temperature, and that in 
this condition it is specifically heavier than in the molten state. 
In reference to these views it may he remarked that the phenomena 
of freezing of water and of the solidifying of molten cast iron are 
not quite suitable for comparison, as the ice formed by the solidifi- 
cation of the water remains at the same temperature \ the iron, on 
the other hand, changes in temperature after solidification, cooling 
gradually until the temperature of the surrounding air is reached. 
It is sometimes argupd that the sharpness of the edges of castings 
from iron and some other metals is due to the assumed expansion on 
setting, but as this sharpness appears in practice to be attainable 
by hot run metal, whilst dull run metal gives blunt or rounded 
edges, it is doubtful how far it can be accepted as proving any 
change due to expansive action. 
Again, it is evident from foundry practice that, so far as the 
volume of the finally cooled solid and the volume when molten is 
concerned, that an allowance of about /g-th part (varying somewhat 
with the size and shape of casting) has to be made in patterns as 
representing the lineal contraction during such cooling, and that in 
consequence we must conclude that the cold solid is specifically 
heavier than the molten metal. Under these circumstances we 
naturally ask why should the solid metal float in metal specifically 
lighter than itself ? It is obvious to any one who has made such 
experiments that the floating of the solid iron cannot be due to any 
supposed phenomena of expansion at setting, as the piece would 
require to be heated up to the white hot condition or critical point 
at which it would just begin to melt. And it is found when the 
floating pieces are taken out and broken through, they only in some 
cases — depending on their size — show a dull redness. 
3. From the author’s experiments already mentioned, and from 
others sinee carried out and to be summarised, it appears that the 
bulky pieces sink at first , reappearing at the surface in longer or 
shorter times according to their weight and form, and that small 
pieces either disappear for only a short time, or remain floating on 
the surface. Indeed, small flat shaped pieces are unsuitable for 
these experiments, as the constantly forming skin on the molten 
