HIS IMPROVED KYK-riKCES. 270 



sharp at the edges, unless the pencil of rays coming from 

 the object in nearly parallel lines, and which enters the 

 eye after having passed through the eye-piece, be suf- 

 ficiently broad. This being once granted, the inference 

 followed, th.it an image ceases to be well defined, when it 

 does not strike at least two of the nervous filaments of the 

 retina with which that organ is supposed to be overspread. 

 These gratuitous circumstances, grafted on each other, 

 vanished in presence of Herschel's observations. After 

 having put himself on his guard against the effects of 

 diffraction, that is to say, against the scattering that light 

 undergoes when it passes the terminal angles of bodies, 

 the illustrious astronomer proved, in 1786, that objects 

 be seen well defined by means of pencils of light 



rhose diameter does not equal five tenths of a milli- 



letre. 



Herschel looked on the almost unanimous opinion of 

 the double lens eye-piece being preferable to the single 



ms eye-piece, as a very injurious prejudice in science, 

 experience proved to him, notwithstanding all the- 



>retic deductions, that with equal magnifying powers, in 

 reflecting telescopes at least (and this restriction is of 

 some consequence), the images were brighter and better 



lefined with single than with double eye-pieces. On one 

 occasion, this latter eye-piece would not show him the 

 bands of Saturn, whilst by the aid of a single lens they 

 were perfectly visible. Herschel said : " The double eye- 

 piece must be left to amateurs and to those who, for 

 some particular object, require a large field of vision." 

 (Philosophical Transactions, 1782, pages 94 and 95.) 



It is not only relative to the comparative merit of 

 single or double eye-pieces that Herschel differs from 

 the general opinions of opticians ; he thinks, moreover, 



