1894.] The Molecular Surface-energy of the Esters, Sfc. 163 



At equal intervals of temperature above absolute zero all normal gases 

 possess equal molecular volume-energy. 



By " molecular volume-energy " is understood the product of pres- 

 sure into molecular volume, that is, into the volume occupied by the 

 molecular weight of the gas taken in grams ; while molecular surface- 

 energy signifies the product of surface-tension and molecular-surface, 

 that is, the surface on which equal numbers of molecules are supposed 

 to be uniformly distributed, equal to the two-thirds power of the 

 molecular volume of the liquid. 



The apparatus employed for low temperatures was that figured in the 

 * Transactions of the Chemical Society,' vol. 63, p. 1094. A double set 

 of observations was made, each set with a different capillary tube. 

 One of the tubes was accidentally broken during the experiments, 

 and was replaced by one of approximately the same radius. The radii, 

 as described in a previous paper, were measured by means of a micro- 

 scope with micrometer eye-piece ; tube A had a radius of O01843 mm. ; 

 tube B of 0'01708 mm. ; and tube C of 0'01046 mm. These measure- 

 ments were confirmed by determining the ascent of pure benzene in 

 the tubes at known temperatures, and this is, on the whole, the easiest 

 and most accurate method of determining their diameters. 



For higher temperatures, the apparatus, described in the ' Philo- 

 sophical Transactions,' 1893, A, p. 662, was employed. In order to 

 apply a correction for the capillary rise in the barometer-tube in 

 which the capillary tube D was confined, a determination was made 

 with each ester at some temperature (usually the boiling point of 

 ilcohol under atmospheric pressure) at which capillary rise had been 

 letermined in a wide tube with tube A, B, or C, where correction was 

 innecessary, the ascent being taken, as customary, in inverse propor- 

 tion to the radii of the tubes. As the variation of capillary rise with 

 jmperature is approximately a linear one, a sufficiently accurate 

 arrection may be obtained by assuming a rectilinear relation. Thus, 

 for example, if at 78 the rise in the wide tube was 80 mm., and in 

 le narrow tube 27 mm., it was necessary to add 3 mm. to the rise in 

 ic narrow tube at that temperature. At the critical temperature 

 the correction is, of course, zero, since at that temperature there is no 

 capillary ascent in any tube. It was held that this difference 

 decreased in the barometer tube proportionately with rise of tempera- 

 ture, so that, for example, if the critical temperature were 278, at 

 the temperature 178 the correction applied amounted, to 1*5 mm. 



We regarded it as unnecessary to increase labour by taking ob- 

 servations at each 10 rise of temperature, since a few points on the 

 curve serve to show whether the rectilinear relation hol'ds. The plan 

 of experiment was as follows : The pressure tube containing the 

 capillary tube was heated ill the vapour of chlorobenzene, boiling 

 under atmospheric pressure about 132, the exact temperature 



li 2 



