1903.] The Ultra-violet Spectrum of Radium. 297 



moved away from the slit the spectrum gains in definition ; the greater 

 the distance between the light and the slit the finer is the sharpness 

 and definition of the lines, but at the same time the loss of light is 

 great. This loss of light, however, may be obviated in great measure, 

 with considerable improvement of the definition, by inserting optical 

 condensers between the spark and the slit. I use two condensers 

 of quartz, piano-cylindrical, one being double the focus of the other, 

 the axes intersecting each other at an angle of 90 ; the object being 

 to concentrate a line instead of a point of light on the slit. All 

 optical condensers waste much light. I use them more to obtain well- 

 defined images than to abridge the time of exposure. 



The slit is made of two shallow-angled quartz prisms, as I have 

 already described,* and the distance apart of the jaws is generally 

 O'Ol mm. 



The quartz prisms, of which there are five, are of 60, made in two 

 halves of 30 each, according to Cornu's plan, one half being right- 

 handed and the other half left-handed. The contact surface of both 

 bisects the refractive angle of the entire prism, and is placed per- 

 pendicularly to the crystalline axis of both prisms. In this way 

 duplication of lines is avoided. 



The sensitive film-holder must be set at an angle varying with the 

 portion of the spectrum being photographed, as the focus of the less 

 refrangible is longer than that of the more refrangible rays. More- 

 over, the focal plane is not flat, so the film itself must follow the 

 diacaustic curve, or the lines on it will not all be in focus together.! 

 For this reason glass plates cannot be used, and celluloid films are 

 employed. 



To obtain the best definition of any desired line for measurement, 

 expose for a long time and develop briefly, using plenty of potassium 

 bromide in the pyro developer. It is impossible to photograph 

 properly the whole spectrum with a single exposure, so as to have it 

 well defined in all parts, since the brightest lines are over exposed 

 and blurred sideways before the faint ones are impressed. This im- 

 portant fact is too generally lost sight of in spectrum photography. 



For correct determination of wave-lengths, it is necessary to photo- 

 graph on the same film the spectrum of a metal whose lines are known. 

 I generally use iron for this purpose ; it has the advantage of giving 

 a large number of very fine lines, the wave-lengths of which have been 

 accurately measured, and not being very volatile, the poles do not 

 rapidly wear away. By means of diaphragms close to the slit the 

 experimental and the standard spectra are photographed on the same 



* 'Chemical News,' vol. 71, p. 175, April 11, 1895; 'Roy. Soc. Proc.,' vol. 65, 

 p. 241. 



f 'Chem. News,' vol. 72, p. 87, August, 1895; and vol. 74, p. 259, November, 

 1896 ; ' Proc. Roy. Soc.,' vol. 65, p. 242. 



