168 CARNEGIE INSTITUTION OF WASHINGTON. 



of half a wave-length of light, are sensibly polarized in a plane normal to the 

 lines of the grating, (b) A diaplu-agm of the rectangular type, for use in the 

 image plane of the eyepiece in order to cut out all light except that from the 

 particular object under examination. The field should, however, cover at 

 least 10°. (c) The importance of a field intensity of illumination approaching 

 that of day-Hght and best adapted for the eye at any particular time. The 

 simplest method for securing this is by means of a substage polarizer in 

 conjunction with the polarizing prism; the polarizer can be rotated, and with 

 it the intensity of illumination of the field varied. These factors are not im- 

 portant for ordinary observations, because the resolving power there required 

 is not great; but in liigh-power, critical work they are significant and enable 

 the observer to accomphsh with comparative ease that which inider other 

 conditions is a matter of difficulty. 



(6) Examination of ores and metals in polarized light. Fred. E. Wright. Mining and 



Met., No. 158, Sect. 9 (1920). 



A discusssion is given in this paper of the different methods which may be 

 used in the determination of opaque minerals by the effect which they have on 

 vertically incident light-waves, either non-polarized or polarized. On reflec- 

 tion from a birefracting, biasorbing crystal plate, such waves are modified to 

 some extent, and these changes can be recognized and measured with suitable 

 apparatus. 



A description of these instruments and of the several factors underlying 

 their use is given. Several new methods are suggested which have proved to 

 be satisfactory in practice. 



(7) A trigonometric computer. Fred. E. Wright. J. Wash. Acad. Sci., 10, 29-31 (1920). 

 The trigonometric computer described in this paper was devised and con- 

 structed in the Geophysical Laboratory for the purpose of computing the ancles 

 in obhque spherical triangles with a degree of precision about one-half minute of 

 arc. This instrument has been in use since 1913; it enables the observer to 

 save an appreciable amount of time in computing the angles of a spherical 

 triangle which otherwise would have to be done by logarithms. 



(8) A graphical method for plotting reciprocals. Fred E. Wright. J. Wash. Acad. Sci., 



10, 185-188 (1920). 



In the search for a mathematical function which shall represent satisfac- 

 torily the data obtained from a series of experiments, it is convenient in certain 

 instances to plot the reciprocals of one of the variables and from the curve 

 thus obtained to deduce the form of the desired equation. In this paper a 

 simple method for accomplishing this purpose is described and the principle 

 explained on which the method is based. 



(9) Estimating impurities by means of the melting-point curve. Walter P. White. J. 



Phys. Chem., 24, 393-416 (1920). 



Freezing-points where the thermometer is immersed in the substance are 

 more reliable and precise than those by the capillary-tube method. If, in 

 addition, the form of the freezing curve is observed, there is obtained an 

 indication of the amount of impurity which is independent of all previous 

 knowledge or uncertainty as to the melting-point of the pure substance, and 

 even of the absolute accuracy of the observer's thermometer. If the deter- 

 minations are a control of a purification process, there are also a saving of time 

 and the avoidance of certain serious chances of error. 



Smallncss of dimensions diminishes local temperature differences and is very 

 often a superior substitute for stirring of the tested substance. It also 

 economizes both time and material. A small thermo-couple or thermel of 

 several couples lends itself very well to use in small test samples. 



