20 OPTICAL PROJECTION 



fracts it to /2: then the distance between F and /2 repre- 

 sents the aberration of the D h combination. But, owing 

 to the curvature, away from the lens D, of the meniscus d, 

 the marginal ray passes through d nearer the centre than 

 through h, and consequently its second refraction by such a 

 lens is less on that account ; the same ray also passes through 

 the meniscus at a less angle of incidence, which in another 

 way also reduces the second refraction. Consequently, the 

 marginal focus is lengthened, and the aberration is reduced 

 to the distance from P to /. 



The main factor in this correction is the bending away 

 from each other at the margins of the two lenses, which is 

 obtained equally in the double piano form, and explains its 



superiority to the two double 

 convex lenses, c, fig. 10. But 

 it will be evident that the other 

 condition, of ' minimum devia- 

 tion ' at the margin, is only ap- 

 proximated to when the curves 

 or thicknesses of the lenses are 

 in some proportion to the foci on 

 each side of the condenser (i.e. 

 the position of the radiant, and 



the position of its image on the other side of the condenser). 

 Hence, for a lime-light condenser, the lens next the radiant 

 should be of considerably deeper curve, the two lenses taking 

 the form of fig. 12 rather than of E in fig. 10. Then the 

 spherical aberration F/ will also be comparatively smaller. 

 A thicker lens, however, is more in danger of cracking from 

 the heat ; therefore, as it will be obvious that a somewhat 

 smaller diameter at d d will collect all the bundle of diverging 

 rays which can reach the second lens, D, this fact should 

 be taken advantage of, in order to reduce its thickness while 

 keeping the deeper curve (see fig. 13). 



All things considered, I regard this as practically the best 



