88 REPORTS ON INVESTIGATIONS AND PROJECTS. 



The earlier measurements of mineral melting-points contain an excellent 

 illustration of this. But a few years ago a long list of minerals was melted 

 and the melting temperatures measured with an apparatus which yielded 

 results correct within perhaps 25 °. Most of the important minerals were 

 found to fall within the region between 1100 and 1300 C. The result 

 obviously merely served to divide the minerals, somewhat roughly, into eight 

 classes upon the basis of melting temperature, a classification so rough as 

 to be of but little service to petrology. Over against this illustration, it may 

 be noted that as soon as attention was properly directed to the situation, 

 improvements were devised and made available for such measurements which 

 now enable us to detect temperature differences smaller than o.i° in the 

 behavior of minerals in this region — a precision entirely comparable with 

 the corresponding measurements of physics at ordinary temperatures. 



The Geophysical Laboratory is making a direct effort to meet just this 

 desideratum, and to make of petrology an exact science also. Throughout 

 the history of this laboratory we have sought to begin with problems which 

 permitted accurate quantitative definition, both of the materials and of the 

 forces which participate in the reactions as they occur, and to use the suc- 

 cessful solution of one such problem as a stone with which to build a broader 

 foundation for the next. It is a working plan which consciously goes out 

 after difficulties, but it is imperative that laboratory study undertaken in the 

 service of the various branches of geology should become commensurate in 

 scope with the geological problems which it seeks to solve. 



To make an accurate estimate of the character of the work of the year 

 just past, and of our plans for the future, it will be necessary to bear con- 

 stantly in mind that the existing methods and generalizations of physics 

 and physical chemistry are rarely adequate in their present scope to reach 

 even the simpler problems offered by the rocks during formation. A con- 

 siderable proportion of the activities of the Geophysical Laboratory, both in 

 the present and in the immediate future, will therefore necessarily be di- 

 rected by the need for greater scope in the physical and physico-chemical 

 generalizations which can be applied to rock formation. 



To be explicit, the efforts so far made are four in number : 

 ( 1 ) To extend the methods of accurate temperature measurement to include 

 the entire field of rock formation. This work, covering the establishment of 

 a fundamental temperature scale and several modes of applying it to the 

 minerals, is now practically completed for the temperature range from o to 

 1550 C, where all known rocks are molten. 



After devising means to determine the temperature at which minerals 

 form, thus providing a temperature scale in terms of which the history of 

 the cooling earth could be expressed, it became necessary (2) to measure the 

 amount of energy involved in certain phases of the formation process. Here, 

 again, the methods of physics (calorimetry) were complicated in technical 

 character and seriously limited in scope and accuracy. Hardly an attempt 

 had been made to determine the quantity of heat involved in chemical reac- 



