44 



Prof. J. H. Poynting on 



1° difference in the temperature was determined by separate 

 experiments to be 9*4 divisions. 



Time, 

 April 30. 



Pressure, in 

 tmospheres. 



Galvanometer 



deflections, in 



divisions. 



Temperature 



of the cooler 



junction. 



Calculated 

 temperature, 

 at -0073 per 

 atmosphere. 



h m 

 12 51 



12 53 



12 57 



1 30 



1 32 



18 



18 



18 



9 



9 



4-5 

 53 

 5-0 

 2-5 

 3-3 



-•48 0. 



-•56 

 -•53 



-•27 

 -•34 



-13 



-•13 



-•13 



-•065 



-•065 



It will be seen, by a comparison of the last two columns, 

 that the lowering here was four or five times that given by the 

 usual formula. I have not thought it necessary to give details 

 of the other results, as I have not yet had time to investigate 

 the causes of failure. I hope to pursue the subject shortly. 



Perhaps the following imaginary experiment may serve as 

 a simple illustration of the last two sections. Suppose two 

 cylinders, one containing ice, the other water at the same tem- 

 perature, to be connected above by a tube through which the 

 vapour can pass, and let them only be in contact with their 

 own vapour. 



At 0°, or rather at +'0073, their vapour-tensions being 

 equal, as soon as the pressure reaches 4*6 millim. then the ice 

 and water will remain unaltered in amount as long as no heat 

 is allowed to pass into or out of the cylinders. If the tempe- 

 rature be kept slightly below 0°, then, since the vapour-ten- 

 sion of water is now greater than that of ice, the water will 

 gradually distil over into the ice-vessel and there condense as 

 ice, the average temperature rising. If the temperature be 

 kept constant, however, the w T hole of the water will in time 

 go over into the ice-vessel. If the temperature be slightly 

 above 0° (supposing it possible still to keep the ice solid), then 

 the ice has the greater vapour-tension and will gradually distil 

 over into the water-vessel, and the average temperature will 

 fall. In time, if the temperature be kept above 0°, the whole 

 of the ice will go over into the water-vessel. 



If, now, the ice and water be subjected to pressure by porous 

 pistons which the water does not wet (the pressure in each 

 cylinder being the same), then, if the temperature be 0°, an 

 increase of pressure will cause more evaporation from the ice 

 than from the water ; that is, the ice will distil over into the 



