Sir John Conroy. 251 



water cell is just as effective for arresting radiant heat in lantern 

 projection as the alum cell which was formerly considered necessary. 



Another memoir, embodying some conclusions of a practical 

 character, was shortly afterwards published in the " Phil. Transactions 

 of the Royal Society" (A, 1889, p. 245), entitled, "Some Observations 

 on the Amount of Light Reflected and Transmitted by Certain Kinds 

 of Glass." It was shown that the amount of light reflected by freshly 

 polished glass varies with the way in which it has been polished, 

 and that, if a perfect surface could be obtained without altering the 

 refractive index of the surface layer, then the amount would be 

 accurately given by Fresnel's formula. Usually, however, the amount 

 differs from that given by the formula to a greater or less extent. 

 The formation of a film of lower refractive index on the glass would 

 account for a defect in the light reflected, and the changes in the 

 amount of light transmitted, and in the angle of polarization, support 

 the view that an excess is to be accounted for by the polishing having 

 increased the optical density of the surface layer. When light passes 

 through a transparent plate, it is obviously diminished by reflection at 

 the two surfaces, and it is further reduced by obstruction within the 

 plate, due to absorption and to scattering, if the plate is not absolutely 

 homogeneous. Conroy shows that for 1 centimetre the loss by absorp- 

 tion amounts to 2'62 per cent, in the case of crown glass, and 1-15 per 

 cent, for flint glass.* 



In April, 1891, Conroy contributed to the "Phil. Magazine" a 

 paper " On the Change in the Absorption Spectrum of Cobalt Glass 

 Produced by Heat." Cobalt glass gives an absorption spectrum con- 

 sisting of three dark bands, in the red, yellow and green. Conroy 

 showed that, on heating the glass, the absorption between the red and 

 yellow bands diminishes, and that the red band moves further towards 

 the less refrangible end of the spectrum. As the glass cools, it 

 recovers its original colour and resumes its original spectrum. Hence 

 the absorption spectrum of cobalt glass varies with the temperature. 



" The Refractive Index of Water at Temperatures between and 

 10 " ("Proc. Roy. Soc.," 1895, p. 2 28) formed the next subject of research. 



* The paper in the " Phil. Trans.," of which the results are here summarised, 

 had its origin in an inquiry conducted by a Committee of the Trinity House into 

 the relative value of gas, oil, and electricity as lighthouse illuminants. For 

 calculating, from the photometric observations, the intensity of lights shown 

 through lenses and lanterns, it was necessary to make some allowance for loss of 

 light due to its reflection and absorption when passing through glass. But the 

 necessary data could not be found, and Sir John Conroy was asked by the present 

 writer to make the needful measurements, which he did conclusively. Practical 

 inquiries often lead incidentally to new scientific truths, such as was in this case 

 the observation that glass of which the surface is very highly polished transmits 

 less light. 



