ii4 SCIENTIFIC APPARATUS. 



prism of the required angle. If it be liquid, a hollow prism of 

 glass or of quartz, filled with the liquid, is used. Care must be 

 taken that the plates forming the sides of such prisms have truly 

 parallel sides, or faces. In all these experiments for the deter- 

 mination of refractive indices the rays should pass through the 

 prism with minimum deviation; viz., the incident and emergent 

 rays should make equal angles with the surfaces. 



Since the two faces of the prism are of necessity inclined to one 

 another, it follows that every ray in its passage must undergo a 

 deviation or turning from its original. course; and it is sometimes 

 desirable to bring back the ray to its original direction, or, techni- 

 cally speaking, to correct the deviation. This is effected by placing 

 beyond the first a second prism, in a reverse position, of a sub- 

 stance having greater deviating and less dispersive power than 

 the first. By putting together a train of such prisms alternately, 

 direct-vision prisms have been constructed; these disperse the 

 light, but for one particular ray, usually one belonging to the 

 yellow part of the spectrum, do not cause it to deviate. This 

 can, of course, be effected accurately only for rays having a par- 

 ticular refrangibility, that is to say, of a particular period of 

 vibration. The same principle, carried still further, is employed 

 in forming compound achromatic prisms and lenses, the object of 

 which is to correct the colours otherwise produced by the disper- 

 sion due to a single lens ; or, more strctly speaking, to contrive 

 a common focal length for all colours. 



Light when of one colour, or monochromatic, consists of rays of 

 one period of vibration, and consequently of one refrangibility only. 

 The spectrum of such light would consist of a single bright line 

 only. A particular beam of light may be due to waves of various but 

 definite periods ; in that case the spectrum will show lines corre- 

 sponding to those lengths only ; it will be what is called a bright 

 line spectrum. Such are the spectra of the metals when vaporised 

 and rendered incandescent by an electric current. Such are the 

 spectra of gases as a general rule. But this subject, if pursued 



