248 P. IF. Brick/man — Failure of Cavities in Crystals 



outside surface where the two pieces join. This was due 

 to failure to orient the two pieces in their natural posi- 

 t ion. The proper orientation was found after a few trials, 

 and after this the chipping almost entirely ceased, even 

 up to the maximum pressure. The mathematical discus- 

 sion shows that plane cross sections of the crystal warp 

 under pressure, and explains therefore, the necessity for 

 observing the original orientation. All later specimens 

 were marked when cut from the original crystal so as to 

 allow the correct orientation, but this precaution was not 

 taken with the first specimen. 



At higher pressures, beginning at about 6000, signs of 

 failure at the interior appeared, increasing in intensity up 

 to final complete rupture. These signs of failure were of 

 two kinds. In the first place, there was a system of fis- 

 sures in both the hollow piece and the cap. In the hollow 

 piece these began as a set of cracks near the inner mouth 

 of the hole, separated by an angular interval of 120°, cor- 

 responding to the symmetry of the crystal, and penetrat- 

 ing a short distance into the body of the crystal at right 

 angles to the surface. At higher pressures these cracks 

 became more numerous and penetrated to a greater depth, 

 inclining toward the axis at the greater depths. This 

 system of cracks had its counterpart in the solid cap, but 

 the penetration here was always to only a slight depth. 

 The cap also showed cleavage in the interior on surfaces 

 at right angles to the axis, a phenomenon for which there 

 was no counterpart in the hollow piece. These systems of 

 cleavage planes are apparently connected in some way 

 with end effects due to imperfect matching together of 

 the two pieces. A mathematical discussion of these end 

 effects is too difficult to be attempted. 



Final failure of the crystal took place in a way entirely 

 unconnected with these cleavage planes. At pressures of 

 7500 and higher minute flakes scaled off the inner surface, 

 leaving it rough. This flaking proceeded at an acceler- 

 ated rate at higher pressures, at 10,500 amounting to an 

 enlargement of 3% of the volume. On the next applica- 

 tion of pressure, to 11,500, however, the flaking off was so 

 extensive that the upper end of the hole was eroded away 

 to three times its original diameter, and the entire hole 

 was tightly packed with a very fine quartz sand. The 

 density of this sand was of course much less than that of 

 the original solid quartz; so that when the hole had 



