48(> Lord Rayleigh's Investigations in Optics. 



from the lens of the radiant point, the primary, and the 

 secondary focus respectively, expressed in inches. 



Fig. 12. 



Fig. 18. 



u=39 \ 



\ u = 40 



v 1 = 28 ] 



) Vl = 40 



r 2 = 94 / 



J r,= lll 



Fig. 14. 



u =40 

 Vl =60 

 v 2 = large 



Fig. 15. 



u = 40 

 n=117 



v 2 =168 



w=180 

 w x = 32 

 v 2 = 38 



Fig. 16. 



In all these cases the line of intersection of the plane of the 

 lens with the screen lies to the right of the diagram. It will 

 be seen that the primary focal line is thin, though curved, 

 when v 1 = u. This is true in general for an equiconvex lens, 

 as may be shown from considerations of symmetry. 



Experimenting on a plano-convex lens held at an obliquity 

 of about 30°, I found that the focal lines were far better 

 formed when the convex side was turned towards parallel rays 

 than when the flat side was so turned. The theory, which I 

 subsequently investigated, is given in the following section. I 

 think that spherical lenses inclined at the most suitable obli- 

 quity might in many cases, perhaps in star-spectroscopes, 

 replace cylindrical lenses. If it were necessary to cause no 

 convergency at all in the secondary plane, a compensating 

 concave lens, held perpendicularly, would be used. 



[To be continued.] 



