362 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1912. 



vation to the supposed molten surface of the earth as it begins to 

 solidify, we have a suggestion of order and system in its separation 

 into so many kinds of rocks. 



Now, it happens that m the recent development of chemistry much 

 attention has been given to the study of solutions of various kmds, 

 and a great body of mformation has been gathered and classified of 

 which our observation upon the freezing of salt water is a simple 

 type. Still more recently, quite lately in fact, it has occurred to 

 many students of the earth that here lies not only the clue but per- 

 haps the key to their great problem. If the individual components 

 which are intimately mixed in solution separate wholly or partially 

 in some regular way upon freezing — and nearly all the solutions 

 which have been studied appear to show such segregation — we have 

 a quantitative system which will probably prove adequate to solve 

 the problem of rock formation, provided only that the experimental 

 difficulties attending the study of molten rock and the complications 

 imposed by the presence of so many component minerals, do not 

 prove prohibitive. This is a very simple statement of the point of 

 view which has led to the expermiental study of rock formation m 

 the laboratory as a natural sequence to statistical study in the field. 



Geophysics therefore does not come as a new science, nor as a 

 restricted subdivision of geology, like physiography or stratigraphy, 

 but rather to introduce into the study of the earth an element of 

 exactness, of quantitative relation. It may mclude physics or chem- 

 istry, biology or crystallography, or physical chemistry, or all of these 

 at need. The distinctive feature of geophysics is not its scope, wliich 

 may well be left to the future, but its quantitative character. The 

 Geophysical Laboratory of the Carnegie Institution of Washington has 

 entered upon some of the investigations suggested by this long pre- 

 liminary study of the earth — the physical properties and conditions 

 of formation of the rocks and minerals. The Department of Terres- 

 trial Magnetism of the same institution has undertaken another — 

 the earth's magnetism; the German geophysical laboratory at Got- 

 tingen a third — the earthquakes — and these will no doubt be followed 

 by others. 



The first efl'ect of calling exact science into consultation upon 

 geologic problems is to introduce a somewhat different viewpoint. 

 It has been our habit to study the minerals and the rocks as we 

 find them to-day, after many of the causes which have had a share 

 in their evolution have ceased to be active, after the fire has gone out. 

 If we attempt to reconstruct in our minds the operations which enter 

 into the formation of an igneous rock or of a body of ore, we must 

 infer them from present appearances and environment. The experi- 

 mental geophysicist, on the other hand, confronting the same problem, 

 says to himself: Can we not construct a miniature volcano in the 



