456 Royal Society. 



¥ 



This gives for -^- at the freezing-point the value of about 1*09 



~dt 



or 1*10, while its value brought out in the earlier part of the 

 present paper by my brother's quantitative calculation was 1*13 ; 

 and so the feature expected shows itself here in Eegnault's results 

 almost in the full extent in which theory shows that it ought to 

 exist. 



Eegnault gives in the same memoir (page 627 and following 

 pages) another Table, one intended chiefly for meteorological pur- 

 poses, and in which the pressures are stated from —10° to +35° 

 for every T ^ of a degree. In this Table the numbers inserted as 

 representing the pressures below the freezing-point are slightly 

 different from the corresponding ones in his general Table already 

 referred to ; and he mentions that this slight discrepance has re- 

 sulted from the fact that the two Tables were formed at different 

 periods, and were not calculated by the same formula ; but he re- 

 marks that the differences are insignificant, as they scarcely amount 

 to '02 millimetre. Here, too, as in the general Table, the feature 

 expected shows itself, though in a diminished degree. By a careful 

 examination of its column of Differences for -^ of a degree, and 

 by making a few small arithmetical adjustments which may be re- 

 garded as amendments in the way of interpolation in that column, 

 I find that, according to the experimental results as they are re- 



dp' 



presented in this Table, the value of L— at the freezing-point would 



dp 



~di 

 come out to be about 1*05 or 1-06. "We have seen by the new cal- 

 culation, based on theory, in the present paper that it ought to be 

 1*13 ; so here the feature is found showing itself in about half 

 the degree in which, according to the new quantitative calculation, 

 it ought to be met with. "When we consider that Eegnault's re- 

 ductions of his experimental results in the making out of curves, 

 formulae, and tables for representing them in the aggregate were, 

 as we have sufficient ground to suppose, carried out under the idea, 

 now proved to be erroneous, of there being, for aqueous vapour, 

 continuity in variations of pressure with variations of temperature 

 past the freezing-point, just as past any other point of temperature, 

 and when we further consider that the quantities with which we are 

 here concerned are indeed very small, it is not surprising that there 

 should have been a tendency to smooth off this feature on the sup- 

 position that any departures of the experimental observations from 

 the course of a continuous or smooth curve were only slight irre- 

 gularities due to experimental errors or imperfections. 



It may now, in conclusion, be remarked that if from experiments 



