152 Mr Pritchard on forming Diamonds 



of its surface. When this is accomplished, a concave tool of 

 cast iron must be formed of the required curve in a lathe, having 

 a small mandril of about /^ths of an inch in diameter, and a 

 velocity of about 60 revolutions per second. The diamond 

 must now be fixed by a strong hard cement (made of equal parts 

 of the best shell-lac and pumice stone powder, carefully melted 

 together without burning,) to a short handle, and held by the 

 fingers against the concave tool while revolving. This tool 

 must be paved by diamond powder, hammered into it by an 

 hardened steel convex punch. When the lens is- uniformly 

 ground all over, very fine sifted diamond-dust, carefully wash- 

 ed in oil, must be applied to another iron concave tool. (I may 

 here remark, that of all the metals which I have used for this 

 purpose, soft cast iron is decidedly to be preferred.) This tool 

 must be supplied with the finest washed powder till the lens is 

 completely polished. During the process of grinding, the stone 

 should be examined by a magnifying lens, to ascertain whether 

 the figure be truly spherical ; for it sometimes will occur that 

 the edges are ground quicker than the centre, and hence it will 

 assume the form of a conoid, and thus be rendered unfit for mi- 

 croscopic purposes. The spherical aberration of a diamond 

 lens is extremely small, and when compared with that of a glass 

 lens the difference is rendered strikingly apparent. This dimi- 

 nution of error in the diamond arises from the enormous refrac- 

 tive power possessed by this brilliant substance, and the conse- 

 quent increase of amplification, with very shallow curves. The 

 longitudinal aberration of a plano-convex diamond lens is only 

 0.955, while that of a glass one of the same figure is 1.166 ; 

 both numbers being enumerated in terms of their thickness, and 

 their convex surfaces exposed to parallel rays. But the indis- 

 tinctness produced by lenses arises chiefly from every mathema- 

 tical point on the surface of an object being spread out into a 

 small circle ; these circles, intermixing with each other, occa- 

 sion a confused view of the object. Now this error must ne- 

 cessarily be in the ratio of the areas of these small circles, which 

 being respectively as the squares of their diameters, the lateral 

 error produced by a diamond lens will be 0.912, while that of 

 a glass lens of like curvature is 2.775 ; but the magnifying 

 power of the diamond lens will be to that of the glass as 8 to 3, 



