544 REPORT — 1901. 



Department II. — Physics. 

 The following Report and Papers were read : — 

 1. Report on Eadiation in a Magnetic Field. — See Reports, p. 39. 



2. Note on a Metlwd of determining Specific Heats of Jletals at Low 

 Temijeratures. By T. G. Bedford, M.A., and C. F. Green, M.A. 



The specific heats of solids at temperatures below 0° C. have hitherto generally 

 been determined by an obvious slight modification of the method of mixtures as 

 generally used for temperatures between 0° and 10U° C 



It was suggested lo us by Mr. E. H. Griffiths that better results might be 

 obtained by adopting a method which can be regarded as analogous with one 

 which has already been used with success for the 0° C. to 100° C. range, viz., that 

 of Joly's steam calorimeter. The metal, whose specific heat is required, having 

 been previously weighed in water kept at 0° C, is cooled to a low temperature and 

 then again immersed in the ice-cold water. The metal, with the coating of ice 

 thus formed on it, is again weighed in the water. From the difierence in the two 

 weights the mass of ice formed is calculated, the density of ice being known, and 

 thence the specific heat of the metal is obtained in terms of the latent heat of 

 ice. 



The experiments have been merely of a preliminary character, but they have 

 served to suggest the following as an appropriate form of apparatus. 



The metal to be investigated should be enclosed in a cylindrical box, and 

 experiments performed first with the box empty and then when it contains the 

 metal. Then if the walls of the box be of sufficient thickness, this differential 

 method would eliminate to a great extent corrections for the suspension wire, 

 for the gain of heat by the metal and the deposition of hoar-frost upon it during its 

 transit through the air, &c. The essential feature of such a box is that its volume 

 should not be altered by opening and closing it. 



The 'cooler' used by us consisted of three coaxial cylinders. The metal 

 experimented upon was suspended in the inner of the three chambers thus formed, 

 the middle chamber contained the cooling agent and the outer chamber formed an 

 air-jacket. 



It appeared that the best method of determining the temperature of the box 

 would be to bring it into direct metallic connexion Avith the thick walls of the 

 inner part of the cooler and to insert a platinum resistance thermometer in the 

 walls. 



In the method briefly sketched above, the accurate determination of the rise in 

 temperature of the water in a calorimeter in experiments by the method of 

 mixtures is replaced by two weighings. These weighings must, liowever, be 

 performed with great accuracy, since the difierence in the observed weights caused 

 by the formation of ice is only yi^th of the weight of ice formed. 



The chief difficulties of the method are : — 



(1) Uncertainty as to the density of the ice owing to the presence of air in the 

 water. 



(2) The fact that the water cannot be stirred during: the process of weighing, 

 and that therefore its temperature begins to rise above 0° C. and the ice gradually 

 melts. 



3. A Neio Gauge for Siinall Pressures. 

 Bij Professor Edward W. Morley and Charles F. Brush. 



The paper describes two forms of gauge for the measurement of small pressures 

 of a gas. It was especially devised in order to measure the pressure of aqueous 

 vapour. For this purpose McLeod's gauge cannot easily be employed, owing 



