714 



PHYSICS, PROGRESS OF, IX 1890. 



surpass them all in accuracy. The utmost limit 

 of accuracy in the setting of a cross-hair on a 

 fine ruled line is -000002 inch, but direct meas- 

 urement of the length of a wave of green light 

 in the spectrum of mercury vapor showed an 

 error of only -0000001. 



Spectroscopy. Dr. E. Pringsheirn (Berlin Phy- 

 sical Society, Feb. 21) remarks, that by experi- 

 ments on the spectrum of burning sodium vapor 

 it is impossible to answer the question, " Does a 

 gas acquire the power of emitting light rays 

 when its temperature is raised?" since chemical 

 action may alter the conditions. He heated 

 metallic sodium in a closed tube, and obtained 

 the usual bright line, but does not consider Kirch- 

 hoff's law proved absolutely for gaseous radia- 

 tion, since the nitrogen in his tubes contained 

 minute traces of oxygen. Prof. S. P. Langley 

 and F. W. Very (" American Journal of Science," 

 August) report careful experiments with the 

 spectroscope and bolometer on the light emitted 

 by the Cuban fire-fly Pyrophorus noctilucus. The 

 insect produces no heat rays, except those iden- 

 tical with its luminous radiation, while in or- 

 dinary industrial methods of lighting by candle, 

 lamp, or gas, 99 per cent, of the energy produced 

 is wasted in heat. Even in the electric light 

 such waste is enormous. The fire-fly light, there- 

 fore, is far more economical than any yet discov- 

 ered, and Prof. Langley sees no reason why we 

 should not one day produce it in our labora- 

 tories. J. R. Rydberg, of the University of 

 Lund, Sweden ("Philosophical Magazine," April) 

 has found that the " long " lines of the spectra of 

 the elements form doublets or triplets, in which 

 the difference of the wave numbers of their cor- 

 responding components is constant for each ele- 

 ment. This rule had already been announced by 

 Hanly for magnesium, zinc, and chlorine. Ryd- 

 berg finds, also, that the components of the 

 doublets form series, of which the terms are 

 functions of the consecutive integers. These 

 series are of three kinds or groups, which he 

 names respectively ' diffuse," " sharp," and 

 " principal," and the first two kinds are divided 

 each into three orders. Series of the same group 

 and those of the same order are related mathe- 

 matically to one another in a way that shows 

 that they all belong to one system. The wave 

 lengths and wave numbers of corresponding 

 lines, as well as the values of certain constants 

 in the mathematical formula giving the relation 

 of corresponding series, are periodic functions 

 of the atomic weight. Joseph S. Ames, of Johns 

 Hopkins University (ibid., July), has deduced 

 similar relations between cadmium and zinc. 

 C. Runge, of Hanover, discussing the method 

 of E. J. F. Love (see " Annual Cyclopaedia," 

 1889, page 095) for discriminating between real 

 and accidental coincidence of lines in spectra, 

 agrees with Love that if the curve representing 

 errors of coincidence diverges from the theoreti- 

 cal error curve the supposed coincidences are dis- 

 proved, but asserts that even if the curves agree 

 there is no proof. He shows that for a certain 

 distance of lines in one spectrum the plotted 

 curve must always resemble the error curve for 

 any lines that one pleases to take as lines of the 

 other spectrum. This conclusion invalidates 

 many of the supposed proofs of Griinwald's 

 theory (" Annual Cyclopaedia," 1889, page 695). 



Interference. P. Garbe (" Journal de Phy- 

 sique," IX. 47), has found that two kinds of 

 bands are produced when light is passed through 

 two similar gratings. The first are bands of 

 which the central one is colored like the others, 

 and the color changes periodically if the gratings 

 be shifted or rotated. If the slit be small the 

 second kind of bands true interference bands 

 are produced. 



O. Wienen (Wiedemann's " Annalen," XL, 

 203) has photographed stationary light waves in 

 a sensitive transparent collodion lamina, whose 

 thickness was only about -y the length of a wave 

 of the light used. This lamina was placed be- 

 tween two glass plates at a small angle with a 

 metal mirror. After exposure, layers were dis- 

 covered in the lamina, which were due to ac- 

 tion at the ventral segments of the stationary 

 waves produced by reflection. The experiment 

 seems to prove conclusively that the chemically 

 active vibrations of polarized light are at right 

 angles to the plane of polarization. Probably 

 the vibrations of the luminiferous medium pro- 

 duce vibrations of the solid particles in the 

 same direction which causes the photochemical 

 change. 



Absorption. Herzberg Schulze, in experi- 

 ments on the absorptive power for light of dif- 

 ferent kinds of glass, finds that thick, heavy 

 glass absorbs 27 per cent. ; less heavy, 12'6 per 

 cent. ; white Rhenish glass, 10 per cent. ; and 

 ordinary mirror glass, 10 per cent. 



Refraction. E. Doumer (Carl's " Repertori- 

 um," 110, 40-42) finds that all salts of the same 

 acid that contain equal amounts of metal have 

 equal molecular refractive power. For instance, 

 the refractive power of MCI is 21-5 ; that of 

 MCI, is 42-8 ; that of S0 4 M a is 42-5, etc. Prof. 

 S. P. Thompson described Bertrand's refrac- 

 tometer before the London Physical Society on 

 March 7. The instrument depends on the total 

 reflection of a hemisphere of glass, 8 millimetres 

 in diameter, at the end of a tube, plane face out- 

 ward, inclined at an angle of 30 with the axis 

 of the tube. One side of the convex surface is 

 illuminated through a piece of ground glass 

 perpendicular to the plane face. The eye-piece 

 is focused on a scale of ^ millimetre in the 

 tube. A film of the liquid to be measured is 

 spread over the plane face of the hemisphere, and 

 the position of the line separating the light part 

 of the field from the dark part is read on the 

 scale. This differs with the 'liquid that is used, 

 and by calibrating the instrument, the refractive 

 index can be read off at once. This refractometer 

 is remarkable for handiness and accuracy. Hu- 

 rion and Mermeret (Carl's " Repertorium," page 

 110, 1187) have measured the refractive index of 

 gold 1'eaf by observing the alteration of phase of 

 transmitted light with Jamin's interference appa- 

 ratus. They found the index equal to -19 for 

 the spectral line C, to '41 for D, to -72 for b. and 

 for -93 for F. Kundt (" Annual Cyclopaedia," 

 1889, page 697) found -38 for red light and -1 for 

 blue. H. Rubens (Berlin Academy of Sciences, 

 July 24) has measured the refractive index of 

 metals by Kundt's prism method. He finds 

 that light, in passing from iron, cobalt, or nickel, 

 to air, begins by following the sine law for small 

 angles, but for larger ones deviates from it ; so 

 that the refractive index should be calculated 



