12 OPTICAL PROJECTION 



made more convergent so as to do this. We thus learn that 

 for a successful projection, especially of apparatus, the position 

 of the condenser relative to both the light and the object, and 

 to the focussing lens, is almost as important as that of the 

 latter ; and may have to be varied materially, according to 

 any of these circumstances. All the parts of the optical 

 system must be so arranged that (1) as many rays as possible 

 may be made to pass through or fall upon the slide or object ; 

 and (2) also pass through the projecting lens. 



But it is further necessary that the rays should not be 

 more converged than will cause them to pass through the lens. 

 For lastly, we can hardly help seeing that the rays proceeding 

 in fig. 7 from the slide s to the projecting lens L, must be 

 considered in a double character. From the focussing or 

 projecting point of view, they must be regarded as bundles of 

 image-forming rays diverging from every point of the slide s 

 to the surface of the lens L. On the other hand, from the 

 illuminating point of view, we must regard them as converging 

 bundles of rays proceeding from the condenser c to the lens L. 

 It is natural to suppose that this dual character might lead to 

 some want of sharpness or definition in the focussed image. 

 In ordinary lantern arrangements, and in exhibiting slides, 

 this is not sensibly the case, because any ray from each point 

 in the slide, to any point on the lens, forms an image on the 

 same spot of the screen ; and thus the luminous rays are 

 identical with the chief image-forming rays. But in movable 

 projecting arrangements, the matter is of the greatest im- 

 portance. Thus, the light might be so drawn back from the 

 condensers as for all the rays to pass through the slide, and 

 then actually converge and cross before entering the focussing 

 lens. 1 All the light would then still pass through the lens, 

 and would reach the screen in some form or other ; but such 

 violent crossing would blurr the image and spoil the even 



1 I have seen this mistake made repeatedly in physical demonstration, 

 especially with the electric light. 



