192 TKANSACTIONS OF THE [mAR. 1, 



at the same time the}' are equally luminous. The shadows dis- 

 appear when the difference of illumination between that por- 

 tion of the screen occupied by them and the adjacent portion is 

 inappreciable and the method evidently depends upon the ability 

 of the eye to detect slight differences in luminosity. 



The Bunsen photometer is the type of a large number of 

 photometers depending upon the disappearance of a spot on 

 the illuminated surface when equality of illumination is ob- 

 tained. The Bunsen screen originally consisted of a disk of 

 white paper having a central translucent grease spot. As 

 originally used, a compensating light was placed at a certain 

 distance behind the screen. One of the two lights to be com- 

 pared was placed in front of the screen at such a distance as 

 to cause fhe central spot to vanish. The distance from the light 

 to the screen having been measured, this light was replaced by 

 the light to be compared with it and the distance adjusted until 

 the spot again vanished. The relative intensities of the two 

 lights can then be calculated from the law of inverse squares. 

 The condition for the disappearance of the spot is, evidently, 

 that that portion of the disk shall be of the same luminosity as 

 the adjacent portions and the accuracy of the measurements de- 

 pends upon the accuracy with which the eye can detect slight 

 differences in luminosit3\ That it is a comparison of luminosi- 

 ties is made very evident if the light on one side of the screen is 

 of a different color from the light on the opposite side. Under 

 this condition the spot can not be made to disappear. 



The polarization photometers of Arago,* Babinet,f and Wild J 

 still further disguise the principle of equal luminosities. The 

 following is a brief outline of the method employed in this class 

 of photometers. 



The light from one of the sources is polarized in a given plane 

 and the light from the other source in a plane perpendicular ta 

 this. The two rays after polarization travel in the same direc- 

 tion, and by viewing the combined ray with suitable apparatus, 

 certain phenomena are observed which disappear when the quan- 

 tities of light polarized in planes perpendicular to each other are 

 equal. With photometers of this class, as with the other classes 

 mentioned, the investigator seeks ultimately to obtain a uniform 

 field, and it is, therefore, evident that the accuracy depends, as 

 in the other photometers considered, upon the accuracy with 

 which the eye can detect slight differences in luminosity. The 

 pala,rization photometers are open to the same objection as the 



• Comptes Rendus, XIII., pages 840, 967. 1841. 

 + Comptes Rendus, XXXVII., page 774. 1863. 

 i Pogg. Ann., XCIX. 



