5G0 wick. 



Difficulties. 



Glass is a very unreliable substance when subjected to high pressure 

 and the breaking of the windows of the apparatus caused considerable 

 trouble. It was difficult to mount the windows in such a way that 

 they should not become cramped against some part of the steel when 

 the pressure was raised. This was especially likely to happen if the 

 ends were not accurately perpendicular to the axis, which was doubt- 

 less the case with some of the windows used since they were made by 

 grinding cylindrical pieces from thick plate glass such as is used in 

 portholes of ocean liners. The curved surfaces were finished off in 

 the lathe and made true but no attempt was made to grind and polish 

 the faces of the windows. The final mounting, shown in Figure 2, 

 seemed reasonably safe. The fibre washer, F, slightly wider than the 

 shoulder, H, was inserted to prevent the crowding of the glass against 

 this shoulder in case the outer face of the window was not exactly 

 perpendicular to the axis. 



A second difficulty arose from the fact that, at high pressures, the 

 rubber of the washers dissolved in the neutral white oil used in the 

 observation chamber. This caused a diminution in the intensity 

 of the transmitted light which took place gradually after the applica- 

 tion of pressure and caused considerable trouble at pressures above 

 2000 atmospheres. When the apparatus was allowed to stand at 

 pressures of this order, the absorption of light continued to increase, 

 a difficulty which proved very serious since the work was done photo- 

 graphically and red-sensitive plates which required long exposures 

 were used. Upon relieving the high pressure after it had been main- 

 tained for some time, the light transmitted through the specimen 

 came back to almost its original brightness at a pressure of approxi- 

 mately 2000 atmospheres. After the pressure had been run up and 

 down several times, the intensity of the light transmitted at normal 

 pressures was somewhat diminished and, upon removal of the speci- 

 men from the observation chamber, the oil next to the windows was 

 found to be cloudy, due to the gradual disintegration of the rubber 

 washers which had to be replaced from time to time to prevent leaks. 

 Different varieties of oil were tried and a heavy variety called " Nujol" 

 was found to cause the least trouble. By changing the oil in the 

 observation chamber frequently the difficulties due to lack of trans- 

 parency in "Nujol" were largely eliminated. 



A comparison of the relative intensities of light transmitted by a 

 given specimen at different pressures is a matter of interest but no 



