108 M. Fraunhofer on the Laws 



this system of lines spectra are produced, which are as large 

 as those produced by very large prisms ; and already in the 

 first spectrum the line D (in the orange) is observed to be as 

 good as double, so that the width of the space can be measured. 

 And since through this system, for instance, D' is already =10® 

 1 4', the law of this modification of light can be deduced with 

 very great accuracy from the experiments made with it. 



In the case when the light fell vertically upon this system, 

 in which 6 = 0,0001223, I obtained, 



These angles are so large, that, in reference to arcs, sines, 

 and tangents, they deviate from each other very materially. 

 Since the instrument with which these angles are measured 

 gives 4" without repeating, some idea may be formed of the 

 degree of accuracy to which they can be depended upon. 



The third spectrum, the fourth, and those following, were 

 indeed well seen with this system ; but the fixed lines in the 

 various colours could not be distinguished with sufficient pre- 

 cision, in order to determine their distances from the axis quite 

 as accurately as those in the first and second. * 



The indicated measures of the first and second spectrum are 



• However great the accuracy may be in reference to the equal distan- 

 ces of the parallel lines, it is natural that it should have its limits, for 

 absolute perfection is unattainable. What influence a slight inequality 

 in the spaces between the parallel lines may have on the distortion or con- 

 fusion of the fixed lines in the different spectra may be judged of from the 



equation 3- ^^^ = ~. For, as it thence follows that cf S-^"^ = ^ J^t—'' 



an inequality which amounts only to the hundredth part of this very small 

 distance •, («r« = 0,00000122 inch,) must produce a disturbance of the fix- 

 ed line D, in the first spectrum of 6' 5", in the second spectrum of 12' 10", 

 in the third of 18' 15", and so forth. This is the cause why, in the first 

 and second spectrum, the fixed lines may still be defined, while in the 

 third and fourth, &c. they are not. 



