X)f Light and its Theory, 109'-. 



in the meanwhile sufficient to lead to a conclusion concerning 

 the law of this modification. 



As the distance of the parallel lines cut upon the glass from 

 each other must be known with the greatest accuracy, and as 

 these lines themselves are difficult to be distinguished by the 

 most powerful microscope, at least cannot be counted, I en- 

 deavoured to etch the first and the last line somewhat stronger 

 than the rest ; and with a microscopic apparatus, adapted for 

 this purpose, I measure the distance from the first line to the 

 last, the etching machine itself reading the lines which are etch- 

 ed. In this way I know accurately how many of them are 

 contained in the etched space. Thus, for instance, the glass 

 of which I have been hitherto speaking contained 3601 lines. 

 Then from the distance of the first from the last, the distance 

 between the middle of any two, that is s, may be very accu- 

 rately known ; and the proportion of the width of a line to 

 the intermediate space between two lines may likewise be 

 very nearly determined. If the lines were so thick that 

 one touched another, and consequently had no space be- 

 tween them, no light could be regularly reflected from the 

 etched surface, and would, as from every other polished sur- 

 face, be dispersed. Were the intermediate spaces equally 

 wide as the lines, the etched surface could only regularly re- 

 flect half as much light as an equal surface of glass that was 

 not etched, therefore the quantity of regularly reflected light 

 from an etched surface of glass is in proportion to the quan- 

 tity of light which is reflected from a surface of glass of the 

 same size not etched, as the width of the spaces between any 

 two neighbouring lines is to the width of those lines. * It is 

 scarcely necessary to remind the reader, that, in experiments 

 thus made, care must be taken that from the second surface 

 of glass no light shall be reflected. ^ 



With another system of line^on glass, in which s =0,0005919 

 of an inch, I obtained with the light falling upon it vertically : 



• Even if the quantity of reflected light could be determined with ex- 

 treme accuracy, this conclusion^ from reasons which shall be hereafter 

 mentioned, could only be considered as an approximation. In wires 

 where « is still much smaller than here, this conclusion would be very ac- 

 curate. In the experiments here mentioned the determination of this 

 proportion is of no consequence, but only that of the magnitude *. 



