PRESSURE ON THE TEMPERATURE OF VOLATILIZATION OF SOLIDS. 47 
increases the extent of its surface by the formation of bubbles. In the case of a 
solid it is obvious that the surface is of limited extent, and it might therefore be 
expected that the solid should rise in temperature. Reasoning thus, the possibility 
of the existence of hot ice was maintained by OaRnelley and other writers in a series 
of letters which appeared in ‘ Nature ’ during the years 1881 and 1882. 
On the other hand, a liquid in the spheroidal state presents a free surface of 
evaporation in every direction, and yet, although exposed to the radiation of a white- 
hot surface, its temperature does not rise to the boiling-point (Balfour Stewart, 
‘Treatise on Heat,’ 3rd edition, p. 124) ; and we find that when water is heated in a 
platinum basin by means of a blowpipe flame impinging on its surface, its temperature 
cannot be raised above 90°. In these cases the surface appears to be large enough to 
allow all vapour to escape with sufficient rapidity to prevent superheating. 
If, then, the rate of evaporation at the surface of a solid is capable of indefinite 
increase, however much heat the solid receives, it follows that solids have definite 
temperatures of volatilization or volatilizing-points, corresponding to definite pres¬ 
sures, as liquids have definite boiling-points. 
Fig. 5. 
A. Condenser from which exit-tube leads to Sprengel’s pump and manometer. 
B. Tube in wdiich thermometer C is suspended, placed in hot bath. 
C. Thermometer, with bulb covered with cotton-wool. 
D. Bulb containing liquid which could be admitted to apparatus by turning 
screw-clip, so as to trickle down the thermometer-stem, and moisten 
the cotton-wool. 
E. Clipped indiarubber tube for admission of air. 
17. It occurred to us that it would be advisable to ascertain if the boiling-points or 
maximum temperature of evaporation of liquids under conditions as nearly as possible 
identical with those to which the solids already mentioned w T ere exposed are the same 
as the temperatures corresponding to their vapour-tensions. To enable this to be 
done, the apparatus was modified as shown in the figure (fig. 5). 
By using a casing of cotton-wool round the thermometer bulb the liquid was 
