Feb. lo, 1 88 1 J 



NA TURE 



343 



to the steam evolved from the lower portions of the ice 

 column being imprisoned and unable to escape, and 

 hence producing pressure sufficient to cause fusion. 



When the greater part of the ice had been melted, the 

 tube was tightly clasped by the hand, the heat of which 

 was sufficient to produce a somewhit violent ebullition. 

 The liquid in boiling splashed up the side of the tube 

 and on to the bulb of the thermometer, where it froze 

 into a solid mass, as represented in Fig. 3. By this means 

 the ice was obtained in moderately thin layers. The tube 

 at the points indicated by the arrows was then strongly 

 heated by the flame of a Bunsen's burner with the follow- 

 ing results :— The ice attached to the sides of the tube at 

 first slightly fused, because the steam evolved from the 

 surface of the ice next the glass, being imprisoned 

 between the latter and the overlying strata of ice, could 

 not escape, and hence produced pressure sufficient to 

 cause fusion, but as soon as a vent-hole had been made 

 fusion ceased, and the whole remained in the solid state, 

 and neither the ice on the sides of the tube nor that on 

 the bulb of the thermometer could be melted, no matter 

 how great the heat applied, the ice merely volatilising 

 without previous melting ; thus proving that if the pres- 

 sure be maintained below the critical pressure the ice 

 cannot be melted. In different experiments the thermo- 

 meter rose to temperatures considerably above the 

 melting- and even the boiling-point of water, the highest 

 temperature reached being iSo' C, when the ice had 

 either wholly volatilised or had become detached from 

 the bulb of the thermometer, but in no case did the ice 

 attached to the thermometer melt when these tempera- 

 tures had been reached, as erroneously stated in some 

 leports of my experiments. The ice attached to the ther- 

 mometer did not partially fuse at thecommencement of the 

 heating, because, the heat reaching the outer surface of 

 the ice first, evaporation could take place from a free sur- 

 face and the vapour not become imprisoned, as was the 

 case with the ice attached to the sides of the tube. These 

 experiments were repeated many times with the same 

 result, except in one case in which the heat applied had 

 been very strong indeed, and the ice attached to the sides 

 of the tube fused completely. On removing the lamp 

 however for a few seconds the water froze again, notwith- 

 standing that the portion of the glass in contact with it 

 was so hot that it could not be touched without burning 

 the hand. 



The chief conditions necessary for success appear to be 

 (i) that the condenser A (Fig. 2) is sufficiently large to 

 maintain a good vacuum. For the size of apparatus 

 given above it ought to be about i litre ; (2) that the ice 

 is not in too great mass, but arranged in thin layers. 

 Nor must it expose too great a surface for evaporation, 

 otherwise the steam is liable to be evolved more quickly 

 than it can be condensed, and the pressure would there- 

 fore rise above the critical pressure. Further, in the case 

 where the heat is applied to the under surface of the 

 layers of ice, the latter must be sufficiently thin to allow 

 of a vent-hole being formed for the escape of the steam 

 coming from below, if not, fusion occurs. When the heat 

 is applied to the free surface of the ice the layers may be 

 much thicker. In order to get the temperature to rise 

 above the ordinary melting-point of ice, it is necessary 

 that a very strong heat be applied, otherwise all the heat 

 is used to convert the ice into steam without raising its 

 temperature ; it must in fact be applied more quickly 

 than it can be absorbed for changing the state of aggre- 

 gation. Prof McLeod, who has written to me to the 

 effect that he has been unable to obtain any symptoms of 

 hot ice, has failed I believe on account of not having 

 complied with this condition. Dr. Lodge, in an admirable 

 and very clear letter to Nature (vol. xxiii. p. 264), has 

 endeavoured to explain why '• hot ice" is possible, and 

 also points out the absolute necessity for supplying the 

 heat more rapidly than it can be absorbed by the vapour. 



Now the question arises, Does the thermometer in the 

 above experiments indicate the real temperature of the 

 ice ? It has been said by Prof Stokes that the ice, though 

 attached to the thermometer, is not at the same tempera- 

 ture as the latter, and that the action is really as follows : 

 The pressure is reduced till the boiling-point falls below 

 the melting-point, and when heat is applied to the ice in 

 contact with the glass tube a film passes into vapour, and 

 thus prevents the ice from touching the glass except at a 

 few isolated points. The great latent heat of evapora- 

 tion prevents the ice from rising to its ordinary melting- 

 point, and hence no fusion occurs. The ice is only heated 

 — except at the few isolated points of contact — by radia- 

 tion,and therefore comparatively slowly. A portion of the 

 heat passes through the ice and falls on the thermometer 

 inside, and the latter rises in temperature ; this causes 

 th: formation of a film of vapour between the ice and the 

 bulb of the thermometer, so that the latter is in contact 

 with the ice at a few points only, and therefore hardly 

 any heat passes by conduction to the ice. 



Asunder the circumstances of the case this appeared 

 the most probable explanation of the phenomena, it was 

 of great importance to show by other and more con- 

 clusive experiments whether the ice really was hot or not. 

 For this purpose Prof. Roscoe suggested the most de- 

 cisive test which could be applied, viz., dropping the 

 supposed hot ice into water and observing the amount of 

 heating or cooling of the latter. Up to the present I 

 have only had the opportunity of completing two of these 

 calorimetrical determinations, and the second of these 

 was merely a qualitative experiment, as the weight of 

 ice dropped in could not be found, owmg to a small 

 quantity of the water having been jerked out of the 

 calorimeter the moment the ice entered it. In both 

 experiments, however, the water distinctly increased in 

 temperature, and therefore showed that the ice must have 

 been above So' C. In the complete experiment the 

 weight of ice dropped into 1S5 grammes of water was i'3 

 grammes, and the rise in temperature o-2^C., showing that 

 the temperature of the ice was 122° C. From the nature 

 of the experiment the weight of ice which could be 

 dropped into the calorimeter was only small, and though 

 the 'rise in temperature was but slight, yet if the ice had 

 been at 0° a relatively large cooling ought to have been 

 observed. Great care was taken to avoid any error in 

 the determinations. The thermometer employed was 

 graduated so as to allow of a difference of o'os" C. being 

 easily detected, two observers read off the temperatures 

 independently of one another, the calorimeter was in- 

 closed in several casings and filled with the water to be 

 used some hours before the experiment, so that it might 

 have the temperature of the room, whilst the time which 

 elapsed between the readings of the thermometer would 

 not be more than about fifteen seconds, and finally the 

 calorimeter was not brought into position to receive the 

 ice till the source of heat had been removed. To place 

 the point beyond doubt, however, several additional and 

 perfectly satisfactory calorimetrical determinations are 

 necessary, and if possible on a larger scale. Such experi- 

 ments are at present in hand. In the meantime I would 

 make t'le following remarks in favour of the high tem- 

 perature of the ice. If the ice is not really hot, notwith- 

 standing that the thermometer indicates say a temperature 

 of 120° C, how is it possible for the ice to hang on to the 

 thermometer? For if it be separated from the bulb by 

 a layer ol" steam, it cannot hang by steam, it would at 

 once become detached from the thermometer. The ther- 

 mometer was chosen so that the bulb was of the same, 

 and in most cases of a less, diameter than the stem, so 

 that there was nothing to prevent the ice falling away if 

 so inclined. 



In some cases I have had thin plates of ice attached 

 by their edge at right angles to the stem of a paper scale 

 thermometer for a considerable time without being 



