2 Lees, Effect of Pressure on Thermal Conductivity. 



difference of temperature. The surfaces of the steel 

 discs and of the discs of material tested were carefully 

 ground on a flat surface so that they would be approxi- 

 mately plane. Thermal contact between surfaces of steel 

 and specimen was improved by smearing both with 

 glycerine and sliding them together. On placing the pile 

 of discs in a small hydraulic press, and applying a 

 pressure equal to that at which the test was to be carried 

 out, any excess of glycerine was forced out and was removed 

 with a cloth. During subsequent decrease or increase of 

 pressure the edge of the film that remained was watched 

 to see whether glycerine was drawn into or forced out of 

 the interval between two discs, but no such motion 

 could be detected, and it has therefore been assumed 

 that the glycerine films remained of constant thickness 

 throughout an experiment. The pressure having been 

 reduced to a few pounds on the square inch or to zero, 

 water was sent through the upper and lower steel 

 discs, and a measured electric current through the coil 

 embedded in the middle disc. The heat generated in the 

 coil raised the temperature of the disc in which it was 

 placed, and in consequence, a flow of heat from the middle 

 to the upper and lower steel discs through the discs of 

 material under test, was established. The temperature 

 of the discs, indicated by 4 thermometers in radial holes in 

 the steel discs close to the flat surfaces, became after some 

 time steady, or nearly so, and observation of temperatures 

 of discs and air surrounding them, and of current were 

 then made and repeated every 5 minutes for an hour. The 

 pressure on the discs was then increased to about Soolbs. 

 per square inch, and after a short time observation 

 taken for another hour. Finally the pressure was 

 decreased to a few lbs. or to zero, and readings taken for 

 a third hour. The discs were then taken apart, one 



