TRANSACTIONS OF SECTION B. 



357 



great masters; the many journeymen investigators may be trusted to utilise them 

 according to their abilities. Having once given his great principles to the 

 world, van't Hoff remained practically a spectator of their development ; but 

 by his single act he provided generations of chemists with useful and profitable 

 fields for their labour. 



The following Papers and Eeport were then read :— 



1. The Diffusion of Gases through Water. By Professor Carl Barus. 



Although relatively few measurements of the diffusion of gases _ through 

 liquids have been made, the subject is one of great chemical interest, inasmuch 

 as with given coefficients of diffusion the virtual viscosity of the medium 

 through which a single molecule of the gas transpires through the intermolecular 

 pores of the liquid may be computed. Hence for a variety of liquids and gases 

 the data should throw light on intermolecular structure. 



The method consisted in finding the temperature-pressure conditions of the 

 flotation, at a given level of a cylindrical cartesian diver, in which the gas to 

 be examined had been imprisoned. It is remarkable both for its extreme sim- 

 plicity and its astonishing accuracy, provided a room of constant temperature 

 is at hand ; for the few milligrams of gas contained may be weighed with the 

 same relative accuracy with which absolute temperature and pressure may be 

 measured. , . 



The finite and differential equations of the phenomenon (the latter being 

 expressible in terms of the density of the diffusing gas, if not simple) were 

 discussed and lantern-slides exhibited showing details of the apparatus the 

 mass-time graphs throughout a period of months for the interdiffusion of air- 

 air, hydrogen-hydrogen, air-hydrogen, hydrogen-air, and oxygen-hydrogen, 

 through water, and the constants for the first two cases. Whereas the graphs 

 for a single gas are linear, those for pairs of gases are of indefinite variety, 

 and the very curious result of a gas apparently diffusing against the obvious 

 pressure gradient is frequently met with. Such anomalous results are explained 

 iu terms of the partial pressures of the constituents of the imprisoned impure gas. 



2. The Present Position of Electric Steel Melting. By Professor Andrew 

 McWilliam, A.R.S.M., M.Met.—See Reports, p. 261. 



3. The Compressibility of Mercury. By Dr. Wm. C McC Lewis. 



It. has been shown recently 1 that the following expression holds within the 

 limit of error for the majority of normal liquids: — 



Tot 



L = ,-»- 



where L=the latest heat of vaporisation of the liquid, T^abs. temp., a- 

 coefficient of expansion of the liquid, /3 = compressibility of the liquid, and 

 p = density of the liquid. , 



In the case of mercury, however, the discrepancy between observed and 

 calculated L is great; and as both L and a are known with considerable 

 accuracy, it seemed likely that the cause of the discrepancy was to be found in 

 the usually accepted value of 13. L is calculated from the vapour-pressure data 

 given by M. Knudsen 2 ; the values of a are those given by Callendar and Moss 

 At 20° C. the calculated value of /? is (1-30 ± 0-02) x 10- 6 per kilo/cm 2 . The 

 most recently determined value of /3, that of P. Bndgman, 4 is 3-70 x 10 on 



• Phil. Mag., July 1911. 2 Ann. d. Physik., 29, 179 (1909). 



8 Proc. Boy. Soc, A. 84, 595, 1911 (Abstract); cf. Tram., i., 1911. 

 o Proc. Amer. Acad., 44, 255 (1909). 



