436 MOSSOTTi ON frauxiiofer's reticular spectra. 



conclusion, that achromatism in dioptric telescopes was impos- 

 sible, which has been long since disproved by experience. 



Following this analogy, Newton prepared a chromatic circle, 

 which was intended to represent the image of the spectrum, inde- 

 pendent of the elongation or contraction which the refraction pro- 

 duces in the different parts of the prismatic spectrum. This, 

 by means of the colours produced by the admixture or super- 

 position of the several component colours, yields very nearly 

 accurate results, but is constructed upon a hypothetical founda- 

 tion. 



Lastly, Newton made use of this same analogy for the forma- 

 tion of a law concerning the places which the different colours 

 occupy in the prismatic spectrum and the length of the corre- 

 sponding fits. This law leads to a remarkable relation which was 

 first discovered by Blanc*, viz. that the length of the fit of any 

 coloured ray is proportional to that power of ^ the exponents of 

 which ai'e obtained when frds of the arc, at the extremity of 

 which the same colour should be placed in Newton's chromatic 

 circle, is divided by the entire cii'cumference of the circle. But 

 the values obtained for the length of the fits or undulations of 

 the different parts of the spectrum according to this proportion, 

 are found towards its exti-emities to differ considerably from the 



truth t. 



2. A better method of ascertaining the composition of natural 

 light and the relation which exists in vacuo or in the air between 

 the length of the undulations of the rays of which it is com- 

 posed and the positions of these rays in the spectrum, consists 

 in the use of spectra obtained by means of a grating, and which 

 were first observed by Fraunhofer. In these spectra, the only 

 element contributing to their formation is the length of the waves 

 of the different rays composing the natural light ; the phaeno- 

 raenon appears in them in its greatest simplicity, without the 

 alterations which the transmission of the rays through a refrac- 

 tive medium produces. Hence in the reticular spectrum (or that 

 formed by gratings) we have a normal spectrum, to which the 

 variable spectra produced in other M'ays may be referred. 



* See Biot, Precis Element, de Plnjs. Exp., edit. 3, vol. ii. p. 434. 



•f- Notwithstanding this critical remark, it is astonishing that Newton, in the 

 first analysis of the spectrum, knew how to combine the different elements con- 

 tributing to its formation by simple and elegant, although only approximative 

 laws. See the note at the conclusion. 



