March 17, 1899.] 



SCIENCE. 



419 



exposed to electrical waves the galvanometer 

 deflection is suddenly reduced to nearly zero ; 

 and when the waves cease the galvanometer 

 deflection is quickly reestablished. This eflTect 

 is described by A. Neugschwender {Wiedemann'' s 

 Annalen, 1S99, No. 2), and the author finds that 

 the film of moisture recovers its electrical con- 

 ductivity so quickly after the cessation of the 

 electrical waves that a telephone in circuit with 

 the silvered strip gives the tone of the induction 

 coil break even when the frequency of the 

 break is very great. W. S. F. 



THE ELECTRIC DISCHARGE IN RAREFIED GAS. 



Mathias Cantor (Wiedemann's Annalen, 

 1899, No. 2), has shown by means of the coherer 

 (a mass of powdered metal forming a portion 

 of an electric circuit), that the electric discharge 

 produced through a vacuum tube by a large 

 storage battery gives off electric waves. This 

 discharge must, therefore, be either oscillatory 

 or intermittent, contrary to the notion which 

 has heretofore prevailed. W. S. F. 



brilliancy of light sources. 



In Wiedemann's Annalen, 1898, No. 13, Mr. 

 P. Jenko gives a curiously roundabout method 

 for the determination of the intrinsic brightness 

 or brilliancy of light sources. Before entering 

 into the details, however, it is necessary — such 

 is the confusion of photometric terminology — 

 to state a few definitions. The brightness of a 

 source here signifies the total amount of light 

 given out by that source and is ordinarily 

 measured in candles. The intensity of illumina- 

 tion of a surface is the amount of light falling 

 upon unit area of the surface and is usually 

 measured in candles per square centimeter. 

 Thus the intensity of the illumination of a sur- 

 face distant one meter from a standard candle 

 (assumed to give off light equally in all direc- 

 tions for the sake of brevity of statement) is 



1 candle 



This intensity of illumination 

 126000 cm^ •' 



is universally but irrationally called the candle- 

 meter. The brilliancy of a light source is the 

 amount of light given off from each unit area 

 of its luminous surface. This, also, is to be ex- 

 pressed in candles per square centimeter. The 

 candle per square centimeter is a convenient 

 unit for expressing brilliancy of light sources, 



but is an inconveniently large unit for exprei^s- 

 ing ordinary intensities of illumination. Thus, 

 easy reading requires about tttstjct candle per 

 square centimeter. 



Instead of determining the brilliancy of a 

 light source by dividing its measured brightness 

 (candle power) by the measured area of its 

 luminous surface, making due allowance, of 

 course, for irregular distribution in so far as this 

 is possible, Mr. Jenko illuminates a screen of 

 known area by a light of measured brightness, 

 distance being measured. The intensity of the 

 illumination of this screen is then known. He 

 then comjjares upon a photometer bar the light 

 given off by this screen with the light given off 

 by the source of which the brilliancy is to be 

 determined. He then measui-es the luminous 

 area of the source and calculates its brilliancy 

 in terms of the brilliancy of the Illuminated 

 screen, using an obvious relation between brill- 

 iancies, brightnesses and distances along the 

 photometer bar ! W. S. F. 



THE magnetization op iron. 



In Wied. Ann., Band 66, No. 13, pp. 859-953, 

 Max Wien communicates the results of a most 

 careful and elaborate investigation upon ' The 

 Magnetization of Iron by Alternating Currents.' 

 The first part of the paper contains a general 

 resume of the literature of the subject, with a 

 useful set of references to the original articles. 

 Following this comes a discussion of the mag- 

 netization of iron by alternating currents, in 

 which it is shown that for a coil containing an 

 iron core and having a purely sinusoidal E. BI. 

 F. applied to it, neither the induction nor the 

 magnetizing force will be a simple sine function 

 of the time, but will contain higher harmonics, 

 on account of the varying permeability of the 

 core, and that also the apparent resistance of 

 such an electro-magnet is greater than the re- 

 sistance of its windings, while its apparent is 

 less than its true self inductance. 



A full description of the experimental ar- 

 rangements and necessary corrections for Fou- 

 cault currents, upper harmonics, etc., is then 

 given together with the values obtained for the 

 induction and hysteresis for irons of various 

 qualities, using magnetizing currents having 

 frequencies of 128, 256 and 520 per second. 



