GEOPHYSICAL LABORATORY. , IO3 



such accuracy is unnecessary, since the variations of density due to inhomo- 

 geneities of the material may be much greater than this. 



Powdering a crystalline substance does not change its density by an 

 amount which we can detect with certainty, provided that the material is 

 homogeneous and free from cracks and holes; but if the substance is not 

 homogeneous, then, as might be expected, the fine powder is denser than the 

 coarse particles. 



Neither does very high hydrostatic pressure produce any after-effect on 

 the density of strictly homogeneous crystalline compounds. 



But if the pressure be not uniform, then the density of a metal which has 

 been subjected to such compression — or has been deformed in any other 

 way — usually increases first (owing presumably to the filling up of pores and 

 cracks) and then decreases, sometimes even so as to reach a final density less 

 than the original value. Subsequent annealing of the specimen causes a 

 renewed increase of density. The direction of the change of density on 

 compressing bismuth is, contrary to Spring's conclusion, the same as that for 

 other metals, namely, a decrease of density following upon deformation. 

 The bearing of these results upon the question of the "flow" of metals is dis- 

 cussed : they are shown to be in harmony with the idea that the "flow" — or 

 indeed any deformation — of a metal is a manifestation of a real melting 

 produced by the unequal strains set up during the process. 



Finally, it is important to emphasize the fact that the density of most sub- 

 stances is somewhat variable, owing to a lack of complete homogeneity of 

 the material. In consequence of this, slight changes of density can not be 

 regarded as good evidence for the occurrence of any transformation or 

 chemical reaction, whether produced by subjecting the system to compres- 

 sion or by other means. 



(11) Die Dichte fester Stoffe, mit besonderer Beriicksichtigung der durch hohe Drucke 



hervorgerufenen dauerenden Anderungen. John Johnston und L. H. Adams. 

 Z. anorg. Chem., 76, 274-302. 1912. 



A German translation of "On the density of solid substances with especial 

 reference to permanent changes produced by high pressures" (J. Am. Chem. 

 Soc, 34, 563-584, 1912). Reviewed under No. 10 above. 



(12) A correlation of the elastic behavior of metals with certain of their physical 



constants. John Johnston. J. Am. Chem. Soc, 34, 788-802. 1912. 



This paper consists of a discussion of the idea that the "flow," or perma- 

 nent distortion, of metals is conditioned by a real melting, not of the whole 

 mass of metal at any one instant, but of successive groups of particles 

 (namely, those on which the brunt of the strain momentarily falls). This 

 idea serves to correlate some properties of metals which at first sight would 

 appear to bear no relation to each other; it leads, namely, to the fact that 

 there is a parallelism between all the elastic properties of metals for which 

 quantitative measurements have been made and the pressure — assumed to act 

 on the solid phase, but not, or not to the same extent, on the liquid phase — 

 which is required to lower the melting-point to ordinary temperature. This 

 pressure is a function of the melting-point, latent heat of melting, and den- 

 sity at the melting-point of the metal; hence, if these quantities are known 

 for any substance, we can predict the relative order of magnitude of any of 

 its properties which imply deformation of the material. 



The same mode of reasoning is equally valid for any crystalline substance, 

 and could be applied to all salts (including silicates and other geologically 

 important substances) if the necessary data were available. At the present 



