162 ANNUAL BEPORT SMITHSONIAN INSTITUTION, 1925 



and of the solid as consisting of the same molecules regularly 

 arranged. 



. The discovery of tlie true character of tlie melting curve is of evi- 

 dent imi)ortance for geophysics. 



In addition to an indefinitely rising melting point, the measure- 

 ments showed that solid and liquid become more and more alike in 

 volume at liigli pressures, but tliat the latent heat absoi-bed on passing 

 from solid to liquid is little changed. It is furthermore universally 

 true that the liquid is more com])ressil)le than the solid. 



This is natural enough for normal licjuids, but is not so obvious 

 in the case of ice, which has a volume larger than that of liquid 

 water, but is less compressible. All these facts must be explained 

 by the future theories, but nothing essentially new is involved here ; 

 when we have an adequate theory of the solid and the liquid state 

 separately, the correct explanation of the melting phenomena will 

 automatically follow. 



Another interesting fact brought out by the measurements on 

 water and some other substances which crj'stallize in two or more 

 modifications is that in the liquid there may be nuclear structures 

 of considerable complexity capable of persisting for days. The 

 liquid is not the simple thing that it may appear to a casual glance, 

 but, at least under some conditions, it may carry concealed Avithin 

 it traces of its past history imperceptible to ordinary means. The 

 existence of these structures in the liquid is now being demon- 

 strated by X rays, but the extraordinar}'^ persistence of the indi- 

 vidual structure can not be shown by such means. 



One minor way in which the pressure tool may be of value is 

 in determining latent heats of melting. A calorimetric measure- 

 ment is admittedly difficult. By measuring the melting pressures 

 corresponding to two melting temperatures near the normal melt- 

 ing point and determining the changes of volume, one has an 

 indirect measurement of latent heat which is often far preferable 

 to a direct measurement. The pressures involved are not high — 

 only a few hundred atmospheres — and the apparatus is so simple 

 that it may be readily constructed. 



Contrasted with the phenomena of melting are those of poly- 

 morphic transitions in the solid state. These are of all degrees 

 of complexity and are governed by a few simple rules. No cases 

 have been found of a critical point between two solid modifications; 

 this doubtless corresponds to some essential physical necessity, for 

 it is hard to see how one kind of space lattice which is characteristic 

 of one crystalline modification can change, without discontinuity, 

 into some other space lattice characteristic of anotlier crystalline 

 modification. Except for this, we may apparently have nearly every 

 kind of behavior. Under varying conditions of pressure and tem- 



