GEOPHYSICAL LABORATORY, 89 



the crystallization of rocks with and without pressure, the solution of one kind 

 of rock in another, and, in short, all the phenomena which concern the trans- 

 formation of magma to crystallized rock and crystallized rock to magma." 



We can perhaps best render an account of our stewardship by reviewing 

 this group of problems as it appears from the laboratory viewpoint after the 

 lapse of eight years. 



It is first of all to be noted that the need of the geologists which is voiced 

 by Professor Van Hise is all for measured facts and accurately defined rela- 

 tions, not for additional theories or mere indications. It obviously aims to 

 make of geology an exact rather than a descriptive science. It is proposed to 

 measure melting temperatures, crystallizing temperatures, specific gravities, 

 conditions of equilibrium, pressures, and in general accurately to define the 

 conditions surrounding rock formation exactly as the physicist and chemist 

 have done in the development of their sciences, which now find universal 

 application. 



When such problems are placed in the hands of the research student who 

 is familiar with the laboratory viewpoint, he at once goes to the heart of the 

 matter with questions like these : Can apparatus be developed which will 

 measure the temperature of liquid rocks without courting its own destruc- 

 tion? Can the rocks* themselves be chemically defined and separated into 

 groups containing one mineral, two minerals, or three minerals each, and no 

 others? Can the proposed pressures be applied at the temperatures where 

 these rocks form? Will it prove practicable to measure specific gravities at 

 extremely high temperatures ; and, if so, what standards of comparison can 

 be employed, etc. ? 



Although at that time none of these things had been very satisfactorily 

 done, the tremendous possibilities of such research made it appear well worth 

 while to risk much in an attempt to attain such powerful instruments of 

 scientific investigation. 



Without entering upon unnecessary detail, the attempt has now been made 

 to control and measure the temperatures of rock formation with appropriate 

 accuracy, to obtain (by making them in the chemical laboratory) rocks* which 

 contain but one or two or three minerals and wholly free from the disturbing 

 impurities commonly found in nature, to make accurate determinations of 

 the density of these mineral and rock types, just before and just after crys- 

 tallization, and finally, to experiment with such rock specimens under accu- 

 rately measured pressures. 



Furthermore, solutions have been formed of one mineral in another under 

 conditions where all the minerals which crystallize out can be carefully 

 studied. In cases where these crystals have proved minute in size, new 

 methods of identification have been developed to enable us to meet that con- 

 dition also. The effect of adding a third component to such a mixture of 

 two has also been studied at all the temperatures where changes occur. 



* The word "rocks" is here used in a somewhat broader sense than the historic 

 geological definition, but the meaning will be readily understood. 



