5^6 Dr Brewster on a New F ejecting Telescope.. 
surface, and also from imperfections of figure, as it; is universally 
admitted tha,t a surface of glass is much superior., to a metallic 
surface; and Newton has himself remarked, that everydrre:^ 
gularity in a reflecting superficies, makes the rays sivdij Jim or 
six tmus more out of their due course than the like irregularis- 
ties in a refracting one 
The construction by w hich I propose to remedy these disad- 
vantages, is shewn in Plate V. Fig. 1. where AB is the spe- 
culum reflecting the parallel rays II A, liB to a focus at F, 
The cone of rays AFB is intercepted by an Achromatic Prism 
GH, which refracts them to foci atyj where a distinct image is 
formed in the anterior focus of the eye-glass E, by which it is 
magnified. The compound prism GH being composed of a 
prism of crown-glass G, and a prism of flint-glass H, united by a 
cement of a mean refractive power, the loss of light sustained by 
the pencil in its transmission through the two, will riot exceed. 
600 rays out of the 10,000, as the light transmitted through a 
lens of glass is, according to Sir William HerscheFs experi- 
ments, 94850 out of the 10,000 incident rays. Hence the 
light lost by transmission through the prism is not one-jjth 
part of the light lost by reflexion, and the errors of reflexion 
arising from defects of surface and of figure are also incompa- 
rably less. As the refracting angle of the prism G will require 
to be larger, in order to produce a given deviation FH^ when 
it is opposed by the refraction of the flint-glass prism H, we 
may place the correcting prism H nearer the focus as shewn 
in Fig. % and make it of crown-glass; or it might even be 
placed at A, beyond, the focus, and in contact with the lens E. 
If this should be found advantageous, the prism and the lens 
might be formed out of one piece of glass, or a single hemi- 
sphere of glass might be . used, the eye-hole being placed at such 
a point of the hemispherical surface, as to obtain a variable 
prism of the required angle united with a plano-convex lens *f*. 
* The deviation of a ray arising from the imperfections of a reflecting surface 
is to the deviation arising from the imperfections of a refracting surface as four to 
one, when the refraction is made from glass into air, and as six to one when the 
refraction is made from air to glass. 
■t Every plano-convex lens, or plano-concave lens, may be converted into a 
lens and a prism combined, by shifting the eye-hole from the vertex of the lens to 
one side. 
