174 Mb KELLAND, ON THE DISPERSION OF LIGHT, 



But since the two errors will be in opposite directions, let us ex- 

 amine the difference between the greatest and least result. 



These differences wiU be found equal to .00016, .00011, .00015, 

 .00013, .00012, .00019, .00012, .00008, .00015, .00020. With only one 

 exception, then, this difference has unity in the fourth place of decimals, 

 and for that exception, which is water, the error is .00020, or barely 

 2 in the fourth place of decimals. 



Results more nearly agreeing might doubtless be obtained by pro- 



ceeding to one place farther in the expansion of sin^-^, but the above 



will suffice to establish the general accuracy of the formula. 



With respect to water, I should have been surprised had the re- 

 sults been more closely coincident, for the values of the refractive in- 

 dices for B and C are respectively 1.330935, 1.331712 from one expe- 

 riment, 1.330977, 1.331709 from another, the difference between these 

 values being in one case .000777, and in the other .000732 ; the former 

 we should have written .0008, in taking to the fourth place only. 

 This difference arises, I suppose, from the different circumstances as 

 regards temperature under which the experiments were performed. 



SECTION III. 



Deductions from the general expression not confined to the explanation of 



Dispersion. 



If our results are founded on correct principles, and are in them- 

 selves right, it is natural to expect that many important and interesting 

 conclusions will follow from them. Some of these I proceed briefly to 

 notice. 



