1^ Mr. J. J. Waterston on Capillarity 



4 . 



In § 6 it wa3 shown that Q= ^; and d being constant, h is 



observed to diminish with the temperature [§ 10], hence 



-^ = —r-. Also the latent heat has been observed to diminish 

 (4 « 



with the temperature, so that SL requires the negative sign ; 

 and the differential equation, which expresses the relation of 

 molecular volume, latent heat, and capillarity of the same liquid 



111- 1 , . ^ . Sm Sh 8h 

 whne bemg acted upon by heat, is — = -r j-. 



It is obvious from this that the first condition required is that 



-T- should exceed t-, because — is positive. The absolute value 

 h L m ' 



of the difference of the decrements of capillarity and latent heat 

 to fulfil the equation is also indicated. 



The most precise observations on the decrement of capillarity 

 arc probably those of M. Wolf [§ 11]. The range of tempera- 

 ture was 25° C, and the middle temperature 13°: the value of 



^ is -0477 : the value of — is -00089 from M. Despretz's ex- 

 h gL '" 



pansion of water, and -p-= •0293 from Regnault's empirical for- 

 mula. Thus we have 



-•0477 --0293 = -0184 instead of -0009. 



The first condition, viz. that -t->- -^, is satisfied, but the abso- 



lute values do not fulfil the equation. We must keep in view 

 that Rcgnault's formula assumes that the latent heat diminishes 

 uniformly with tlie temperature. I have not seen the details of 

 the experiments upon which it is founclecl, but it is unlikely that 

 they can have indicated more than an average result. 



At 5° C. the point of maximum density of water — = and 



SL Bh '" 



y- =-T-. It appears certain from iM. "Wolf's observations, that 



the decrement of capillarity is very uniform, and in tubes of not 

 less than xj^th of an inch diameter is not sensibly different from 

 the amount put down. Between plates the decrement is probably 

 less, on account of the abnormal feature in the capillarity of water 

 in tubes; but it cannot be less than two-thirds of the above, or 

 •0318. The evidence of an increase of the decrement of latent 

 heat and of the specific heat at high temperatures has been stated 

 [§ 15] ; there is consequently no reason to suppose that the 



ST 



value of -p- can be greater than "0293, unless the peculiarity in 

 the expansion of water aflfects it iu an abnormal way. 



