154 CARNEGIE INSTITUTION OF WASHINGTON. 



As compared with others, the present method is specially advantageous for 

 observations of great absolute precision and wherever it is desirable to secure 

 the advantages which the thermo-electric system possesses in the way of 

 rapidity and of facility in making varied observations. 



(38) The calculation and comparison of mineral analyses. C. E. Van Orstrand and Fred. E. 



Wright. J. Wash. Acad. Sci., 4, 514-525 (1914). 



In this paper the nature of the errors in a chemical analysis is discussed in 

 detail. General mathematical equations are given for the least-square adjust- 

 ment of the data of a mineral analysis; these equations express in concise 

 form all possible adjustments dependent upon averages. The importance 

 of assigning or preferably of determining the weights of the observed data is 

 emphasized. A method for the detection of the systematic errors of an 

 anah'sis is given and illustrated by examples. The various different methods 

 which have been suggested for the adjustment of chemical analyses are con- 

 sidered in the light of these equations and the significance of each method is 

 thereby ascertained. The general conclusion is reached that for most purposes 

 the established method of direct comparison of the weight percentages of 

 chemical analyses is sufficient. 



(39) A simple method for the accurate measurement of relative strain in glass. Fred. 



Eugene Wright. J. Wash. Acad. Sci., 4, 594-598 (1914). 



The method consists essentially in measuring between crossed nicols the 

 path-difference of plane polarized light monochromatic waves from an intense 

 mercury source (546/xm) on glass cubes of uniform thickness. The path-differ- 

 ence is read off directly on a graduated double combination quartz wedge 

 described in 1908 by the writer. With a wedge ground especially to show 

 slight path-differences it is possible to detect and measure path-differences of 

 less than Vm- This degree of precision is sufficient for all practical purposes. 

 The conversion of path-differences into the corresponding mechanical units 

 which produce them is obtained by direct experiment with the particular glass 

 under investigation. 



(40) A new crystal-grinding goniometer. Fred. Eugene Wright. J. Wash. Acad. Sci., 5, 



35-41 (1915). 



Accurately oriented crystal plates and faces are often desirable in crystallo- 

 graphic and optical work. To meet this need several different crystal-grinding 

 goniometers have been constructed. The present precision instrument was 

 built in the workshop of the Geophysical Laboratory, and with it oriented 

 crystal plates can be ground and polished within 1' of the required orien- 

 tation. The crystal can be oriented either optically or crystallographi- 

 cally. In the design and construction of the grinding goniometer special care 

 has been taken to produce a mechanically rigid and precise instrument. The 

 grinding goniometer has been in constant use for several months and has 

 proved to be convenient, accurate, and well adapted for the particular purposes 

 for which it is intended, namely, to grind crystal faces on crystals w'hich are to 

 be measured at high temperatures and pressures; also to grind oriented crystal 

 plates and prisms for optical measurements. 



(41) Spring deposits at Sulphur Springs, Arkansas. C. E. Siebenthal. Microscopic study 



by H. E. Menvin. J. Gcol., 23, No. 1 (1915). 



Analyses of water and sediments from springs, and microscopic study of 

 sediments, indicate that sulphides of iron, zinc, copper, and lead are carried 

 in solution, and that iron sulphides, especially, are being deposited in both 

 amorphous and crystalline states where the waters come to the surface. The 

 bearing of these observations upon the genesis of the sulphide deposits in the 

 Mississippi Valley is discussed. 



