reason. In the latter, the incHnation of the great mirror to 

 the incident rays must deform the image,* and it is now 

 known, that even with faint objects sharp definition is of 

 high importance. It should, in fact, be a segment of a 

 paraboloid, exterior to the axis ; and though a theorem of 

 Sir William Hamilton (Trans. R. Irish Acad., vol. xv. p. 

 97,) might seem to indicate mechanical means of approxi- 

 mating to the figure, yet Dr. R. fears there would be greater 

 difficulty in applying them than in enlarging the aperture of 

 the Newtonian, so as to make up for the loss of light. 

 Another serious objection is, that in the Herschelian the 

 observer's position at the mouth of the tube, must cause 

 currents of heated air, which will materially interfere with 

 sharpness of definition. 



As to the loss of light by the second reflexion. Dr. R. 

 thinks it has been much overrated, and expresses a wish that 

 a careful set of experiments were made on reflexion by plane 

 specula at various incidences, on prisms of total reflexion, 

 and the achromatic prism, proposed as a substitute by Sir 

 David Brewster. 



As to the rest of the instrument, it may suffice to say, 

 that it bears a general resemblance to that ofRamage, but 

 that the tube, gallery, and vertical axis of the stand are 

 counterpoised, so that one man can easily work it, notwith- 

 standing its enormous bulk. The specula, when not in use, 

 are preserved from moisture or acid vapours, by connecting 

 their boxes with chambers containing quickhme, which is 

 occasionally renewed. This arrangement, (which also oc- 

 curred to Dr. R., and has been for several years applied by 



* Any one who has a Newtonian telescope can verify this, by inclining a little 

 the great mirror, so however as not to pass the edge of the plane mirror by the 

 pencil. In Lord O.'s instrument, an inclination of 1 1' sensibly injures it ; were it 

 Herschelian, the inclination must be 3" 11'. 



