Liquid Spheres — Molecular Diameters. 883 



T and p which are independent of variable conditions 

 of temperature, that is, for those values which they have 

 when the temperature is 0° A. But, as Hammick points 

 out, the data for these are not available, or at least are very 

 incomplete. The value of L, the latent heat at the boiling- 

 point, which is equivalent to the work done in dissociating 

 the molecules from the liquid, that is, in creating the free 

 surface of the molecules of the gas and storing it as potential 

 surface energy, may be taken as constant under all con- 

 di'ions after our determinations have been freed from all 

 external work. 



Now, it has been shown experimentally that the surface 

 tension about large masses of any liquid is a linear function 

 of the absolute temperature which holds very approximately 

 for all temperatures of the liquid from its melting-point to 

 the critical temperature 6 C . This result is expressed in the 

 empirical formula 



T, = A + B<9, 



where A and B are constants for any particular liquid. For 

 water the ordinary tables give T = O when # = # c =638° A, 

 and T = 73*3 as the best available value when # = 288° A. 

 From these we have 



T, = 133-6 --209(9, 



so that T =133-6. 



It is evident, however, that only in large spheres or masses 

 can the tension of the envelope obey the law above, for only 

 in these can the heat motion of the molecules next the 

 envelope affect the tension. For spheres containing only a 

 few molecules and for the free molecules themselves an 

 altogether different condition exists, for the heat motion 

 then is the motion of the sphere itself. The tension, there- 

 fore, of the envelope about the free molecule will be very 

 approximately the tension about a large mass when the 

 molecules are at rest, that is, the molecular tension T m 

 will be T . 



If, then, the law holds through the process of solidification 

 and down to 0° A, we. should have for the free water mole- 

 cule T m = 133'6. This seems to be a reasonable supposition 

 when we remember that the forces which cause solidification 

 are in the interior of the molecule or mass and would not, 

 therefore, affect the tension of the envelope on the exterior 

 of the molecule or mass. Moreover (see note at thi* end), 

 there are experimental indications that the surfaces of solids 



