OF THE UNITED STATES. 383 



able in observing. A well defined imixs^e in a niilkish white 

 li2;ht will be found more aij;reeable. 'rhongh the artists say 

 that they can polish an object y;lass as much as they clioose, 

 without detriment to the shape of tlie ti;lass, and therefore 

 accuracy of the image, I should |)refer, from experience, the 

 greater certainty of accuracy with less polish. 



The object glassi-s frequently contain crystallisations, 

 which injure the brii^htness and accuracy of the image. 

 These crystallisations become very apparent in the following 

 mode of trial, which I found to be tiic best forjudging of tlie 

 accuracy of the image : — Wlicn the telescoi)c is adjusted and 

 directed upon a fixed star, the eye part is drawn out, which 

 will cause the microscope which the eye glasses form, to 

 view in its focus successively other sections of tlie cone of 

 light of tiie object glass, and in proportion as the light will 

 be equally diHused upon the circle so seen, and as it will 

 bear to be more drawn out, without becoming dirtused, the 

 telescope will be more perfect and more accurately adjusted. 

 Few telescopes will stand a very scrupulous trial of this 

 kind. .Most generally the light will be much stronger to- 

 wards the circumference of the circle than in the interior, 

 and in this will almost always be found some dark sj)ots 

 of angular shape, |)resenting themselves exactly like crystals 

 under a microscope, which show the cftectof the crystals in 

 the glass, throwing the light from its regular course. Any 

 stray lisiht. as the opticians call the rays wliicli appear as if 

 darting out of tlie circle, is very api)arent in this trial. The 

 best telescopes therefore are tliose which, directed upon a 

 small and not twinklinE: star and put a little out of focus, 

 will present a sharply defined image with a disk like a pla- 

 net. The larger tliis imaije can lie made, the better tht; le- 

 lescope. 



The opticians meet with great difTiculty in finding t:;ood 

 gla><scs: and in makinoc larc;e ol»)ect 'glasses, tnucli more glass 

 is wa.Htcd thai is iH'd. Various other ciicimstanees also 

 increase the price of the larger glasses ; so that a good olject 



