[ 72 ] 

 IX. Intelligence and Miscellaneous Articles. 



ON THE DETERMINATION OF THE ILLUMINATING-POWER OF THE 

 SIMPLE RADIATIONS. BY MM. A. CROTA AND LAGARDE. 

 /~\XE of the most delicate problems of photometry is the mea- 

 " surement of the illuminating-power of lights of different tints. 

 One of us has already indicated* how this question might be solved, 

 by employing a spectrophotometer. That solution supposes that 

 the coefficients of illumination of the different simple radiations 

 composing the light which is to be compared are known. 



If the determination of the radiant energy of a vibration of a 

 determined wave-length can be expressed exactly in thermal or 

 mechanical units, that of its illuminating-power admits of all the 

 uncertainties inherent to the measurement of physiological sensa- 

 tions variable with each eye. There exists no known relation be- 

 tween these two quantities; we have therefore to commence by 

 seeking out what are, for a determined eye, the illuminating- 

 powers of the various simple radiations of the normal spectrum of 

 two sources of light — the sun, and the Carcel standard. 



The best-known treatise on this subject is that of F raunhofert 

 on the Sun ; the results obtained by him are not very concordant. 

 The following is, briefly, the method which we have followed. 



The illuminating-power of a simple light may be regarded as the 

 property possessed by the latter of rendering distinguishable, upon 

 a white screen illumined by it, minute details (lines, characters) ; 

 it will be possible to measure them approximately, as several 

 physicists have done, by weakening this light until the characters 

 can no longer be distinguished, and taking the ratio of the initial 

 intensity to this limit of intensity. The absolute value of these 

 numbers will vary with the fineness of the characters ; but their 

 ratio will be sensibly constant, and will depend only on the wave- 

 length of the light examined. 



The light to be studied (sun or Carcel standard) is received per- 

 pendicularly upon the slit of a spectrophotometer covered with a 

 strip of glass on which is photographed a series of very fine and 

 very close dividing-lines ; the direction of these lines cuts the slit 

 normally. A pure spectrum is then seen furrowed by a consider- 

 able number of very fine longitudinal stria?. If the ocular slit of 

 the telescope be brought onto a region of the spectrum, the simple 

 radiations comprised between two very close known Hmits are 

 isolated ; and by a suitable rotation of the nicol, their intensity is 

 weakened until the striae cease to be perceptible. The phenomenon 

 of the disappearance of the striae is more delicate than one would 

 have at first been inclined to believe ; by practice one arrives at 

 being certain about the degree or the fraction^ of a degree, ac- 

 cording to the region of the spectrum. 



1. We have traced the curve of the wave-lengths as functions 

 of the divisions of the micrometer, and calculated its equation by 



* Oomptes Rendu*, t. xciii. p. 512 ; Phil. Mag Dec. 1881, xii. p. 445. 



t Gilbert's Annalen, xxvi. p. 297 (1817). 



