Radiation of Low Refrangibility through Ebonite. 467 



opaque matter. An exposure of three minutes was given; the 

 exposed half of the slit was then closed and the other half 

 opened, and a spectrum taken through a solution of bichromate 

 of potash T 2 ^ inch thickness. This bichromate was used to 

 prevent the too energetic action of the more refrangible rays, 

 which illuminated the prism and would have caused a veil over 

 the plate. Half a minute's exposure was given. The plate, 

 on development, revealed that rays of very low refrangibility 

 had passed through the ebonite, commencing at W.L. 12,000 

 and extending as far as W.L. 7500; the point of maximum in- 

 tensity was situated at about 9000. The photographs were on 

 a small scale, but sufficed to show the absorption of the ebonite. 

 On the next day we had intended to repeat the experiments 

 with two or three prisms; but the wind had shifted, and 

 the solar spectrum was absorbed as far as about 9000, 

 showing the presence of aqueous vapour. It was therefore 

 useless to experiment further with the sun as a source of radia- 

 tion ; so we used the water of the positive pole of an electric 

 light as a source. It will be seen that the spectrum through 

 ebonite extends to about W.L. 15,000, and then terminates. 



The next point to determine was as to the quality of the 

 beam coming through the ebonite. This we determined as 

 follows — first by placing a piece of ebonite in contact with 

 the photographic plate and throwing an image of the points 

 on it, and thus getting an impression, and then, by a simple 

 arrangement, removing the ebonite to a distance of 1 foot, and 

 allowing the beam to traverse it, and securing another image 

 on a different plate. The photographs showed that the rays are 

 very much scattered in their passage through the ebonite, no 

 distinct image being formed in the latter case, though it was 

 sharp and defined in the former. The amount of scattering 

 it seemed desirable to know. For this purpose the collimator 

 of the spectroscope was used and no prism, the image of the 

 slit 2 7 o mcn w id e was focused on the focusing-screen of a 

 camera, and a piece of ebonite was placed in contact with the 

 plate, and exposure made. This piece was removed and another 

 piece inserted 2 \ inches in front of the plate, and another ex- 

 posure given. The diffusion was most marked: a line-^ inch 

 broad was diffused over a space £ of an inch, most intense, of 

 course, in the centre. By subsequent experiment it was shown 

 that an exposure of three times the length of that given in 

 the first case was necessary to cause the central portion of the 

 band in the second case to correspond in intensity with that 

 of the image of the slit in the first case. With two pieces of 

 ebonite in contact with the plate six times the exposure was 

 required to give the same intensity as with only one plate of 

 ebonite intervening. Hence we may say that the coefficient 



