16 MM. E. du Bois-Reymond and W. Beetz on the 



cally, that the thickness of the ring is in inverse propor- 

 tion to the cube of the radii; in the second table, for the 

 sake of comparison, a calculation is given of the length of the 

 undulations, as found from my admeasurements, but following 

 Becquerel's principle, that the thicknesses are inversely pro- 

 portional to the first powers of the radii. The difference is 

 so great, that in the first series of experiments, in which the 

 radius of the line B was = 13'"*1, that for the line F would 

 be =18'"-1 instead of 14/"*7, to allow of 0-000486 being ob- 

 tained as the required length of the undulations, a difference 

 which cannot in any way be attributed to an error of ob- 

 servation. 



To allow of our comparing the different systems of rings 

 with each other, a question must first be answered which M. 

 Becquerel considered as already determined. Thus, if the 

 index of refraction of the peroxide of lead lies between that of 

 the subjacent metallic plate and that of the air, the incident 

 ray in each case is reflected from a denser into a rarer medium, 

 and thus the minima of the effects of the light when mono- 

 chromatic illumination is used, will occur when the thicknesses 

 of the reflecting films are as the even numbers; if however 

 the film of peroxide of lead is more highly refractive than the 

 metallic plate, the relative thicknesses are as the uneven 

 numbers. I have not succeeded in finding the peroxide of 

 lead among those substances, the refractive indices of which 

 have been given by any experimenters ; and as M. Becquerel 

 does not state whether he was led to the assumption that the 

 refractive index of the peroxide of lead lies between that of 

 the air and the metal, by earlier experiments, his own admea- 

 surements, or mere suppositions, and as such an assumption 

 can by no means be made a priori, I endeavoured to determine 

 this value myself. A film of peroxide of lead was precipi- 

 tated upon a girl plate of silver until it had assumed a uni- 

 form brown colour. It was then capable of taking a tolerable 

 polish. The substance polarized a ray of light reflected from 

 its surface, so nearly rectilinearly, that the angle of polariza- 

 tion could be determined with tolerable accuracy. To control 

 it, a beam polarized at an azimuth of 45° was made to fall 

 upon such a surface of peroxide of lead, the reflected beam 

 was again reflected by a second surface of the same kind, and 

 the angle of incidence at which the elliptic polarization most 

 completely passed into the linear sought for. Both methods 

 gave very nearly the same result, i. e. the mean for the angle 

 of incidence at which the maximum of polarization occurs 

 was 63°. Hence the refractive index of the peroxide of lead 

 is =tang 63°=1 , 964', which is nearly equal to that of the 



