86 THE ROYAL SOCIETY OF CANADA 



The effect of change of temperature is then considered both 

 upon the value of Q and the value of L. Waterston experimented 

 upon liquids in critical temperature tubes of the Cagniard de la 

 Tour pattern, and shewed that the meniscus flattened out many 

 degrees below the point of transition and became convex on fuither 

 heating, 



6 



From m = — he deduces 

 (JL, 



ôm _ ôL ôQ 



m L Q 



4 

 and from the capillary tube equation Q= — 



ah 



SQ _ _ àh 



'Q ~ ~h 



bm __ bh ôL 

 whence — — — ~ — 

 m h L 



and he attempts to fill in thesp temperature coefficients from the 

 experiments of M. Wolf, M. Despretz and M. Regnault with, however, 



ôm 

 only partial success. Waterston takes — as the expansion of the 



liquid per unit volume. (I think here he left out a term which should 

 have been considered, namely the change of density with temperature) . 

 We should say that 



6 dm dQ dL 



m= — gives — - — tt - v 

 QL ° m Q L 



4 dQ dh dp 



and from Q = tt TT^ 1 



dhp Q h p 



dm _ dh dp dL 



whence ' ~ -\ zr 



m h p L 



dm , dp 



or since — , being a coefficient of expansion, = — — , 



dm dh dL 



2 — = ;- , which fits Waterston's data much better.) 



m h L 



"The product mQL has the same constant value for all liquids 

 at any temperature, hence this relation — -assuming it to be proven 

 — enables us to compute the latent heat from the capillarity and vice 

 versa." 



