434 JOSEPH BARRELL 



the molecules, taken as equivalent to their spheres of influence, are 

 in actual contact and suffer mutual compression owing to the 

 attraction of cohesion. The influence of heat is relatively unim- 

 portant in determining the density of a solid. The atoms are in 

 most cases even more compressed and distorted by the converging 

 force of chemical afi&nity than are the molecules by cohesion. This 

 corresponds with the fact that substances of small atomic volume 

 are on the whole more incompressible than those of greater 

 atomic volume. For the more incompressible substances also 

 the decrease in compressibility with added load is relatively little, 

 suggesting that they are already greatly compressed by the forces 

 of chemical affinity.^ The effect of heat serves only to distend 

 slightly the spheres of influence of the atoms, so long as the sub- 

 stance is in the solid state. Rise of temperature to near the 

 melting-point, as long as there is not a softening by the develop- 

 ment of incipient liquidity, should, according to these views, 

 change the elastic properties but slightly. The greatly lessened 

 strength of ice near the melting-point, as expressed in the' freedom 

 of regelation, is not, following these ideas, connected with the 

 slightly lessened incompressibiUty and rigidity. For many sub- 

 stances the problem is complicated, however, by changes of molecu- 

 lar state with changes in temperature and pressure. This is 

 especially true of ice when subjected to extreme ranges in tempera- 

 ture and pressure, as has been shown by Bridgeman; for ordinary 

 glacial ice, however, we deal with but a single state. 



For ice at a temperature of — 7?o3 C. the compressibility has 

 been determined for pressures ranging approximately between 

 loo and 500 atmospheres. It is found to possess, according to 

 Richards and Speyers, about one-fourth of the compressibility of 

 water at neighboring temperatures and about five times the com- 

 pressibility of glass.^ But glass possesses a compressibility between 

 that of acidic and basic holocrystalline igneous rocks. Ice may be 

 taken then as about three or four times as compressible as granite. 



^ "The Present Aspect of the Hypothesis of Compressible Atoms," Am. Chem. Soc. 

 Jour., XXXVI (19 14), 2417-39. 



^ T. W. Richards and C. L. Speyers, "The Compressibility of Ice," Am. Chem. Soc. 

 Jour., XXXVI (1914), 491-94. 



