28 On a Diffusion Photometer. 



of the fatigue attending the use of apparatus which constrain 

 us to shift our attention from one image to another. 



Recently I have changed the nature of the material to the 

 more durable one of glass, of very similar translucency. It 

 may be said that the nearer the glass approaches the uniform 

 soft translucency of paraffin the better it is for the purpose. 

 The durability secured in the use of this material and the 

 accuracy of surface it permits in grinding and polishing are 

 great advantages. I have also heightened the effect of the 

 plane of separation by interposing between the parallelepipeds 

 a film of silver leaf. This is caused to adhere by means of 

 Canada balsam, by which too they are finally cemented 

 together. Thus prepared, their behaviour to light is very 

 striking. In a plane of luminous equilibrium they appear as a 

 single body lit up internally and emitting a soft light. The least 

 flicker in one of the sources of light destroys the illusion of 

 uniformity, and a screen interposed at one side seemingly 

 reduces the parallelepiped to half its size. 



The dimensions I find most suitable are 20 x 50 X 11 mm. 

 for each parallelepiped. They are laid together on two of the 

 larger faces, the parallel external faces being ground smooth, 

 but left unpolished. The surface under observation during 

 experiment is ground smooth and polished after the joining 

 of the parallelepipeds. The most important points to be 

 attended to in their construction are fineness of divisional line 

 and uniformity in thickness. Should there be any difference 

 in the translucency of the parallelepipeds a check observation 

 might be made by turning over the photometer so that the 

 halves change places relatively to the lights, taking a mean of 

 the results. This has not been necessary with such prisms as 

 have come under my notice. 



In judging of the final position of equilibrium I am 

 accustomed to magnify the line with a lens, but I find many 

 observers prefer the absence of the lens. To shelter the 

 photometer from disturbing effects of reflexions, &c, I use a 

 blackened box, somewhat similar in form to that used with 

 the Bunsen photometer, but smaller in size. The usual pre- 

 caution of protecting the eyes of the observer from the 

 direct rays of the lights should, of course, be observed. 



The effect upon which this photometer is based will be 

 extensively observed in natural objects, as translucent minerals, 

 where any discontinuity of substance, as a cleavage crack, 

 occurs. 



A note on the above appeared in the Proc. Roy. Dublin Soc. 

 (vol. iv. p. 345, Dec. 1884), and the photometer was exhibited 

 before the Berlin Physical Society by Dr. Konig (' Nature/ 

 xxxiv. p. 48). 



