636 BRIDGMAN. 



ance, and the discontinuity should be at different pressures at different 

 temperatures. The metals for which such discontinuities should be 

 expected between 0° and 100° are Cd, Pb, Cu, Zn; and Bi, according 

 to Cohen. No such effects were found. The sensitiveness of the 

 measurements may be estimated from the data already given; the 

 accuracy is in most cases great enough so that a discontinuity of the 

 order of 1/100% of the total resistance could have been detected. 

 According to Janecke, however, the transitions all occur above 100°, 

 and none should have been found under pressure, if the phase stable 

 at the higher temperature has the greater volume. I did find a dis- 

 continuity at 140° for antimony, which is much nearer the value of 

 Janecke than of Cohen. This has already been discussed. One of 

 the metals examined is certainly known to have a transition in the 

 temperature range 0° to 100°, tin at 20°. But the transition never 

 starts under ordinary conditions, and one need not expect to find it 

 under pressure. This has already been made the subject of a special 

 investigation.^^ 



In criticism of the results, one mav well admire the skill and care 

 which Cohen and his pupils have bestowed on the measurement of 

 these very minute effects; the existence of the discontinuities which 

 they have found may doubtless be accepted. But it seems to me 

 that their interpretation of the results may well be questioned; the 

 existence of a discontinuity, more or less indefinite, need not of itself 

 be an indication of true polymorphism. It seems that there are 

 many possibilities in the rearrangement of crystalline grains or growth 

 of the larger crystals at the expense of smaller ones (such as have been 

 found by Ewing and Rosenhain ^^ to be stimulated by strains), and 

 that these possibilities of explanation should first be exhausted. The 

 facts that different observers find different transition points and that 

 it is in almost every case necessary to assume more than two modifi- 

 cations to explain the results lend color to suspicion. Furthermore, 

 if the discontiriuities are truly polymorphic in character, there was 

 the best possible chance to detect them under pressure, but none were 

 found. Before the interpretation assigned by Cohen to the results 

 can carry conviction, it would seem to me that we have a right to ask 

 for reproducible results with large individual crystals. It would be 

 worth much effort to prepare such crystals in order to settle this 

 vexed question. 



27 p. W. Bridgman, Proc. Amer. Acad. 52, 164 (1916). 



28 J. A. Ewing and W. Rosenhain, Phil. Trans. (A), 353 (1900). 



