vf Light and its Theot-^j. 105 



7+6 



The numerator in these general expressions for every dis- 

 tinct coloured ray, though a different absolute number, yet, 

 however different the cases may be. is an invariable number, 

 which, as will be easily perceived, relates here to a distinct 

 and absolute measure, the Parisian inch. If this number is ge- 

 nerally marked, for every coloured ray, with w, and the angle 

 of deviation of one and the same coloured ray in the first 

 spectrum with %\ in the second with a% in the third with %"'y 

 and so forth, there is generally 



y '\- 7+0 



And consequently, if v stands for the number which indi- 

 cates to what spectrum in the order the value belongs, (since 

 vis for the axis =0, for the first spectrum = 1, for the 

 second — % &c. and can never be a fraction ;) and if, for the 

 sake of shortness, the sum of the breadth of an interval and 

 a wire, or 7 + 5, == e, then we have generally 



(I-) 



4")- 



The results of the above-mentioned experiments, and also 

 the common expression (I.) thence derived, show that the an- 

 gles of deviation of the same coloured rays, in the series of 

 spectra as they deviate from their axis through the wires, are 

 as the numbers, 0, 1, 2, 3, &c. But the experiments from 

 which these results are derived, gave angles so small, that their 

 sines, tangents, and arcs, are nearly in the same proportion. 

 In my first system of wires, where s = 0,001952 of an inch 

 D^is = 38' 19,3". Upon reflection it will probably appear 

 that in larger angles, that is with finer systems of wires, not 

 the arcs, but perhaps one of the trigonometrical lines of 



