proceedings: philosophical society 103 



interest in a large number of geophysical problems but, owing to the 

 difficulties that offer themselves in the experimental determination, 

 practicalh' no reliable data are available. This paper described a 

 method by means of which the volume-change under hydrostatic 

 pressure of any solid may be determined with an accuracy of about 

 one jjart in a hundred million of the original volume of the solid. Re- 

 sults were presented for the metals gold, copper, silver, aluminum, 

 zinc, tin, cadmium, lead, and bismuth; for the alloys brass, cast-iron, 

 and tin-bismuth eutectic; the minerals halite, quartz, orthoclase, 

 labradorite, oligoclase, pyrite, mica, and enstatite. The pressure 

 range was from 2000 to 12,000 megabars (1 megabar = 0.987 atmo- 

 sphere) . 



In carrying out the determination the solid, surrounded by a liquid 

 such as kerosene, was enclosed in a thick-walled steel bomb fitted with 

 a movable, leak-proof piston and pairs of simultaneous readings were 

 taken of (1) the displacement of the piston and (2) the pressure. 



The P-AV graphs were found to be nearly straight lines, but there 

 exists a slight though distinct curvature such that the graphs are con- 

 cave to the pressure axis. For the more compressible substances, the 

 curvature is suffi<3ient to allow of a rough estimate of the change of 

 compressibility with pressure. This change amounts to as much as 

 10 per cent of the value at thg initial pressure for the most compres- 

 sible substances. 



The average compressibility of the earth at the surface was calcu- 

 lated to be 1.63 parts per million per megabar. 



Discussion: The paper was discussed by Messrs. Sw^ann, White, 

 and Burgess. 



A paper by Messrs. N. S. Osborne and M. S. Van Dusen upon 

 Latent and specific heats of ammonia was presented by Mr. M. S. Van 

 Dusen. The paper was illustrated by lantern slides. 



Using a calorimeter of the aneroid type, specially designed for the 

 peculiar conditions, the specific heat and latent heat of vaporization 

 of ammonia have been determined throughout the temperature interval 

 — 40° to + 40°C. The apparatus used has been previousl}' described 

 in detail before the Society. It consists essentially of a cylindrical 

 metal shell suspended within a thermally controlled metal jacket. 

 The metal shell or calorimeter, containing the ammonia to be investi- 

 gated, is provided with an electrical heating coil and a platinum resis- 

 tance thermometer. 



In the measurements of specific heat two independent methods 

 were used, each of which avoids sources of error present in the other. 

 In the first method, the heat added to a fixed amount confined in the 

 calorimeter under saturation conditions and the resulting change in 

 temperature are measured By using data for the specific volumes of 

 the two phases and the latent heat of vaporization, the corrections for 

 vapor are applied, giving the specific heat of the saturated liquid. 



In the second method the calorimeter is kept full of liquid at a con- 

 stant pressure. The heat added to the variable amount in the calorim- 



fe 



