Apbil 5, 1907] 



SCIENCE 



531 



to the plates, the outer an oxidation 

 product giving the brilliant surface colors 

 which are due to a thin film. Sometimes 

 there is another layer with an exceptionally 

 brilliant silver-like surface. 



The Absorption of Some Solids for Light 

 of Short Wave-length: Theodore Ly- 

 man. 



The grating spectroscope which the au- 

 thor has used in his measurement of 

 the hydrogen spectrum between 1,650 

 and 1,250 Angstrom units, though well 

 adapted for the determination of wave- 

 lengths, is not especially fitted for rapid 

 work in other branches of the subject. A 

 prism spectroscope would be preferable for 

 many purposes on account of its short light 

 path and the superior intensity of its 

 spectrum. 



The first purpose, therefore, of the pres- 

 ent research was a purely practical one, 

 namely, to find some substance from which 

 the prism and lenses of such an instrument 

 might be constructed. White fluorite is 

 the only known substance fitted for the 

 purpose, and it was hoped that a less costly 

 substitute might be discovered. 



In order to test rapidly the transparency 

 of a great number of substances an in- 

 strument of special type was constructed. 

 In it the light from a discharge tube fell 

 upon a concave mirror and was thrown by 

 it through a fluorite prism on to a screen 

 coated with vnllimite. The whole appa- 

 ratus was enclosed in an air-tight case 

 which could be exhausted and filled with 

 hydrogen. The specimens to be examined 

 were introduced into this case in such a 

 manner that eight of them could be inter- 

 posed in succession between the mirror and 

 source of light without opening the appa- 

 ratus. The extent and intensity of the 

 spectrum on the fluorescent screen was ob- 

 served through a glass window. 



The substances examined were: Colored 



fluorites, quartz, topaz, gypsum, eelestite, 

 rock salt; barite aliun, colemanite, sugar 

 (rock candy), borax adularia, calcite, chry- 

 soberyl, sanidin, anagonite, and apophyl- 

 lite. The results are: 



I. No substance shows so great trans- 

 parency as white fluorite. 



II. The transparency of colored fluorites 

 varies through a considerable range, but 

 some light green specimens are nearly as 

 good as the white variety. 



III. Colored fluorites may be deprived 

 of their color by heating. The process 

 does not materially alter their trans- 

 parency. 



IV. Quartz in thicknesses of about 1 

 mm. is transparent to k 1,500. The absorp- 

 tion increases with such rapidity with the 

 thickness that prisms and lenses would cut 

 the spectrum, off at about A 1,750 for all 

 practical purposes. 



V. Of the remaining substances Ceylon 

 topaz is the best, being transparent to 

 A 1,560. 



VI. Rock salt is transparent to A 1,750 

 only. 



Geometrical Theory of Radiating Surfaces 

 with Discussion of Light Tubes: Edward 

 P. Hyde. 



Theoretical photometry assumes two gen- 

 eral laws of radiation. (1) The law of 

 variation of the intensity of illumination 

 of a surface in face in inverse proportion 

 to the square of the distance of the surface 

 from the luminous source is merely a state- 

 ment of a geometrical property, if the 

 rectilinear propagation of light is assumed. 

 (2) Lambert's law of variation of the in- 

 tensity of a luminous surface in direct 

 proportion to the cosine of the angle of 

 emission is an empirical law based pri- 

 marily on the observation that a uniformly 

 bright sphere, when viewed at a distance, 

 appears as a uniformly bright disk. It 

 would seem to follow from Kirchhoff 's law 



