514 VAN ORSTRAND AND WRIGHT: MINERAL ANALYSES 



trough. The test showed that if the object (say muscle, or nerve) 

 evolved heat at the rate of 1 X 10"^ g-cal. sec.-^ it can be detected. 

 The appendix consists of notes on (1) galvanometer mirrors, 

 (2) vacuum galvanometers, (3) the most efficient combination 

 of thermopile and galvanometer resistance, and (4) the m.ain- 

 tenance of high vacua by means of metallic calcium. The cal- 

 cimn is contained in a quartz glass tube which is attached to the 

 vacuum (stellar) thermopile. The thermopile container is ex- 

 hausted by means of an oil pump, and the stopcock is closed. 

 Thereafter the vapors given off are removed by heating the 

 metallic calcium with an alcohol lamp. The device is thoroughly 

 reliable, as evidenced by the fact that it was carried to the Lick 

 Observatory, Mt. Hamilton, California, and back, the vacuum 

 having been maintained for over two months with no signs of 

 failure of the calcium. 



CHEMISTRY. — The calculation and comparison of mineral, 

 analyses. C. E. Van Orstrand, Geological Survey ^ and 

 Fred. E. Wright, Geophysical Laboratory. 



Two different methods have recently been suggested^ for 

 the calculation and comparison of mineral analyses and each 

 method is considered by its proposer to be the best method 

 available for the purpose. Now the term "best" depends on 

 the criteria which are used in comparing the different methods 

 and before proper decision can be made in such instances, it is 

 essential that the criteria themselves be examined in detail 

 and with particular reference to their competency and bearing 

 on the final result. The present problem has to do with data of 

 observation and is therefore subject to the criteria and. methods 

 in current use in the exact sciences for the discussion and compari- 

 son of such data. 



It will be of interest in the following pages to treat the prob- 

 lem from this viewpoint and. to determine the exact relations 



1 Published with the permission of the Director of the U. S. Geological Survey. 



2 W. T. Schaller, The calculation of mineral formulas. J. Wash. Acad. Sci., 

 3: 97-98. 1913; R. C. Wells, The interpretation of mineral analyses. J. Wash. 

 Acad. Sci., 3:416-423. 1913. 



