246 P. JT. Bridymom — Failure of Cavities in Crystals 



This ensured that the stress conditions in a single uncut 

 cylinder with an axial hole were reproduced as exactly as 

 possible. The orientation of the two parts with respect to 

 the original crystal was marked, and the two pieces were 

 always, except at the very first, fitted together in the 

 original orientation. This is an important point to which 

 I will return. The outside surfaces of the cylinders 

 were very nearly circular, and w r ere polished. The inner 

 holes were nearly but not quite coaxial with the outer 

 surface, and were nearly but not quite round. During the 

 test the inner hole contained only air at atmospheric pres- 

 sure, except for a small device to be described later. The 

 outer ends of the cylinder were ground flat, but were not 

 polished. 



A piece of soft rubber tubing was slipped over the out- 

 side of the cylinders and was tied to brass end pieces 

 as shown in fig 3. Between the brass end pieces and the 

 crystal was a thin piece of hard rubber or red fiber, to 

 equalize any slight inequalities in the surface of the brass. 

 The combination was then immersed in a liquid in a heavy 

 steel cylinder and stress applied by producing hydrostatic 

 pressure to any desired amount in the liquid. The func- 

 tion of the soft rubber tubing was to freely transmit pres- 

 sure to the specimen and at the same time to keep the 

 liquid from the surface of separation of the two cylinders. 

 It is evident that under these conditions the entire outer 

 surface of the cylinders is exposed to the same hydrostatic- 

 pressure. 



The cylinder in which pressure was produced was of 

 chrome-nickel steel, 8 inches outside diameter and 1V S 

 'inches inside diameter. Pressure could be raised to 

 12,000 kg/cm 2 with this apparatus. The details of con- 

 struction and methods of measuring pressure have been 

 previously described. 2 A smaller cylinder in which a few 

 tests were made permitted a maximum of 24,000 kg/cm 2 . 



The following measurements were made. The outside 

 diameter of each piece at the two ends and the middle and 

 at angular intervals of 30° was measured to 0-0001 inch 

 with a micrometer. Six measurements of the outside 

 length were also made with the same micrometer. The 

 denth of the hole was measured with a depth gauge to 

 0-0001 inch; this measurement was alwavs somewhat 

 unsatisfactory because the bottom of the hole might be 



2 P. W. Bridgman, Proc. Amer. Acad., 49, 626-643, 1914. 



