ADVISORY COMMITTKK ON GEOPHYSICS 29 



ers, while the questions in each will stimulate and react helpfully 

 upon one another. These are (i) the great envelopes of the eaj-th , the 

 atmosphere and the hydrosphere, which constitute the chief sources 

 of external activities and condition the habitability of the globe ; 

 (2) the body of the eaj-th, whose crust records its history, and whose 

 interior is full of dark, intricate problems ; and (3) the niotio7is and 

 external relations of the earth, which condition its form, its external 

 temperature, its illumination, and probably its magnetism and other 

 important phenomena. These lie in the common ground of several 

 sciences, notably in the over-lapping border-fields of physics, chem- 

 istry, geology and astronomy, but they present special problems 

 whose central interest is terrestrial. 



The only hope of adequate solution of these profound problems 

 lies in special experimentation. Neither the application of existing 

 science, with its limitations, nor of pure theory, with its present 

 narrow basis, covers, in any competent way, the ground of these 

 problems. Their magnitude and complexity are such as to require 

 a degree of extension of deductions quite beyond the limits war- 

 ranted by present experimental data. It is therefore necessary that 

 special experiments shall be devised to determine with the utmost 

 precision the laws of variation, with varying conditions, under as 

 great a range of mass, pressure, temperature and time as possible, 

 and then to apply these determinations with the most critical cir- 

 cumspection to the phenomena which the earth itself presents, and 

 to check them by cross-investigation based upon independent 

 sources. The soul of the method should be to determine by the 

 grandest and at the same time the most refined artificial experi- 

 ments what is the meaning of the magnificent experiments expressed 

 in the evolution of the earth. The experimentation should there- 

 fore be shaped with special reference to the phenomena of the earth, ^ 

 and be checked by all the cross-lines of evidence that can be drawn 

 from geological and astronomical data. 



{c) Some of the problems to be investigated. It is quite beyond 

 the scope of this report to set forth adequately the particular ques- 

 tions to be investigated, but the mention of some of the more 

 declared ones may serve to give definiteness to the undertaking 

 recommended. 



I . The envelopes of the earth doubtless in large measure had a 

 common origin, and they are still most intimately related in action 

 and function. In a narrow sense the atmosphere is the domain of 

 meteorology, but its greater problems reach back into the domains 



