500 Scientific Intelligence. 



physiological chemists. The physical chemistry is so clearly 

 presented that the book may be recommended to chemists as 

 well. h. w. FOOTE. 



6. Spectral Determination of Temperature. — Mr. Irwin G. 

 Priest has recently described in the Bureau of Standards Sci- 

 entific Paper, No. 443, a method by which the temperature of 

 an artificial light source may be determined by the process of 

 color matching. The light from a chosen source is passed 

 through a quartz plate between nicols, and so arranged that 

 the spectral distribution of the emergent energy for all needed 

 positions of the second nicol shall be that characteristic of a 

 complete radiator, or the so-called black body. As the tempera- 

 ture of the latter may be calculated by Planck's formula any 

 source which emits light which is capable of evoking a color of 

 the same hue and saturation as the Planckian radiator may be 

 regarded as having the same temperature. As to the precision 

 of color comparisons, quite apart from this particular method, 

 it was found that when the color fields of two lamps were 

 matched by varying the voltage of one of them, the probable 

 error of a single observation, when translated into temperature, 

 was about 6° at 2850° absolute. 



With the apparatus as designed, the change of color by 

 rotation of the nicol was equivalent to the variation of black 

 body temperature so that one scale can be calibrated in terms 

 of the other. Various tests of the reliability of the method 

 were made. It was found e. g. that a certain 500 watt lamp 

 had a temperature of 3086 °K as determined radiometrically, 

 that is, by the spectral energy distribution, but was measured 

 as 3082 °K by the color temperature apparatus. In the case of 

 a 900 watt stereopticon lamp the two methods gave respectively 

 3087° and 3085°. The color matching method was also applied 

 to the estimation of the temperature of the crater of a 65 volt 

 10 ampere carbon arc for which the author found mean values 

 of 3780 °K with solid carbons, and 3420 °K for cored carbons, 

 with errors possibly in the neighborhood of 50°. 



A widely different method for the estimation of the tempera- 

 ture of a source is reported by R. T. Birge in the May number 

 of the Astrophysical Journal for the current year, as a result of 

 the study of the variation in intensity of certain series lines 

 in band spectra. By means of the recent formulation of the law 

 of such a series, it is possible to express a relation between the 

 temperature of the source, the convergence separation of the 

 lines, and the parameter of the line of the series for which the 

 intensity is a maximum. From a study of the spectrogram of 

 a carbon arc using 13 amperes at 170 volts this author estimates 

 the temperature of the vapor to be 4500°±300°K. The same 

 method applied to the reversing layer of the sun indicates a 

 temperature of 4000° ± 500K. f. e. b. 



7. Vector Calculus; by James Bryne Shaw. Pp. V, 314. 



