356 On a neiv Species of coloured Fringes 



When the fringes on the second image were inclined to the? 

 right, those on the first image were inchned to the left; so that, 

 both in point of position and magnitude, the two sets of fringes 

 follow a different law. 



The preceding measures of the magnitude of the fringes at 

 different obliquities, were not taken wicli that accuracy which is 

 necessary for determining the law of their variation. I have 

 made numerous experiments ibr this purpose ; but when the 

 angle of incidence is considerable, there is always such a degree 

 of distortion in tlie fringes, and such a perceptible variation in 

 their magnitude, from the slightest cliange in the position ot the 

 eye, tliat I found it quite nnpracticable to take measures in 

 which any confidence could be placed. This difficulty no doubt 

 arises from the Imperfect flatness of the surfaces of the plates 

 of glass ; and I fear that even our best artists are not capable of 

 producing better plates than those which I used in the jjreceding 

 experiments. The following measures may be considered as to- 

 lerably correct. 



Tlie inclination of the plates was not the same as in the pre- 

 ceding experiments: 



A ng!-es of Incidence. Number 

 acre 



36 56 



58 48 



62 52 



71 30 



if the two parallel plates are placed at any dislance la/iatevcr, 

 and the preceding experiments repeated, the fringes will be 

 found to suffer no change either in tlieir magnitude or direc- 

 tion. 



1 now took three plates of parallel glass, that gave tlie co- 

 loured fringes when any two of them were put into the proper 

 position. When the third plate was placed either before the 

 other two, or between them, or behind them, it did not in the 

 least degree affect tlie fringes which they produce!'!. When xt 

 was placed, however, in such a position as to form a new re- 

 flected image, this image vvas also crossed by the coloured 

 fringes. 



Whe the third jjiecc of parallel glass was cemented with 

 Canada balsam upon the face of the Jirst plate, or upon the back 

 of the second plate, the fringes disappeared. When the inter- 

 val between the tw3 plates was filled with water, or with Ca- 

 nada balsam, the fringes were very faint, though distinctly per- 

 ceptible. Hence it follows, that the production oj the fringes 



depends 



