GEOPHYSICAL LABORATORY. 89 



tions at temperatures as high as those at which minerals form, and even in 

 the most accessible temperature region, that is, at the temperatures of every- 

 day life, an error of 0.5 per cent was usually considered inevitable in such 

 measurements. Six papers have been published from this laboratory during 

 the present year which deal with the problems of calorimetric measurement 

 alone. Of the result attained, it may be said that specific heats can now be 

 measured in any portion of the temperature range stated above with an accu- 

 racy ten times greater than that usually found in similar measurements at 

 ordinary temperatures. It remains to devise methods of applying this system 

 to the determination of heats of formation as well. 



(3) The third direction in which the available methods proved inadequate 

 was in the microscopic examination, or more generally in the determination 

 of the crystallographic and optical character of the crystals formed in our 

 solutions. Early in our investigations it appeared that the attempt to pro- 

 duce in the laboratory crystal types of high chemical purity almost invariably 

 resulted in crystals so small, when compared with those found in nature, that 

 the usual methods of study were not equal to the task unless considerable 

 sacrifices of accuracy were made. From our quantitative viewpoint, obvi- 

 ously, sacrifices of this kind were carrying us in the wrong direction. The 

 question of methods and technique in the determination of crystals under 

 the microscope was therefore taken up with similar thoroughness and is now 

 practically completed. Submicroscopic crystals, of course, are no more ac- 

 cessible to study than heretofore ; but crystals of but a few hundredths of a 

 millimeter in their chief dimensions can now be determined with as great 

 certainty as is ordinarily possible in large natural crystals. Not only this, 

 but the precision attained in technical measurements of this kind has made 

 necessary a fundamental study of the optical principles underlying the pass- 

 age of light through inactive plates which sets definite limits to the attainable 

 accuracy in a certain class of optical measurements and is a contribution of 

 considerable value to the general subject of physical optics. The paper re- 

 ferred to is very briefly reviewed on page 95 and the entire collection of new 

 optical methods will be published in book form during the coming year. 



(4) The fourth task which the limitations of existing physical methods 

 has imposed upon these studies is the total inadequacy of methods for the 

 application and measurement of high pressures, especially when combined 

 with the high temperatures necessary for many of the problems of rock 

 formation. In this direction, however, we are not yet ready to offer more 

 than a general report of progress on account of the very serious technical 

 difficulties which have been encountered. 



A fifth difficulty can also be foreseen in the limitations of physico-chemical 

 theory and experience with which to attack polycomponent systems. Combi- 

 nations of two minerals or of three find more or less helpful analogues in 

 existing studies of salt solutions at low temperatures, but when four or more 

 minerals are found together, as in the case of most of our rocks, their rela- 

 tion must be determined ab initio. The methods which have been thus far 



