282 OPTICAL PROJECTION 



it with water. A trough of this kind is very convenient for 

 absorption experiments also. 



That lenses bend rays of light in the same way may be 

 shown, by passing the same beam of light through the edge 

 of a large lens, and also by the experiments described in 

 Chapter I. 



160. Dispersion. With a glass prism of 60, or a pris- 

 matic bottle filled with carbon disulphide, colour phenomena 

 become conspicuous, while the deviation of the rays is far 

 greater. The simplest arrangement for projecting the spec- 

 trum is shown in fig. 149. A slit (from 1 to 3 mm. wide) is 

 placed in the optical front, and focussed on the screen. The 

 prism is placed just beyond where the rays cross, when the 

 image of the slit will be greatly turned aside, and converted 

 into a spectrum. 



Note on Deflected Projections. Fig. 149 only gives the 

 essential apparatus. The lantern may be turned aside on a 

 bare table, and the prism-stand adjusted thereon in its place. 

 But in this class of experiments, which include also reflected 

 rays (as those from a soap -film later on), it is on the whole most 

 convenient to employ such an arrangement as fig. 99, where 

 the slit is first focussed on the screen, then the prism placed 

 in position, and finally the whole arrangement, including 

 the lantern, rotated until the spectrum comes again upon 

 the screen. In this particular case all the apparatus would 

 still be in the optic axis of the lantern ; but often a focussing 

 lens has to be placed on one side of this, e.g. to focus the 

 soap-film (fig. 179, p. 327). In this case the sliding cross-piece 

 c D, of fig. 99, affords all the necessary accommodation. 



Another way of producing the spectrum is to project parallel 

 rays through a vertical slit on the open front of the nozzle, the 

 objective being removed ; and to focus it with the loose lens. 

 Which plan is adopted will often depend upon the manner in 

 which the lantern was left arranged from any preceding ex- 

 periment a consideration which will often vary precise details. 



