contributions to the Spectral Analysis of the Stars. 363 



we shall succeed not only in detecting the influence of the earth's 

 motion, but in determining its amount with such accuracy as is 

 desirable for a preliminary control of theoretical conclusions. 



The numbers adduced signify parts of the micrometer-screw, 

 and refer to the distance of the two sodium-lines : — 



Mean 



Sodium-flame. 

 49*5 

 50-5 

 53-0 

 49-5 

 . 50-6 + 0-6 



Mean 



Sun. 

 49*5 

 51-5 

 48-1 

 48-9 

 49-6 + 0-5 



In the following series of observations the reversion-spectro- 

 scope had been provided with another micrometer-screw with a 

 somewhat coarser thread, and also two other systems of prisms, 

 the dispersion of which in the vicinity of the sodium-line is 1*77 

 as much as that of the system used for the above measurements. 

 In this case, also, the former achromatic object-glasses of the col- 

 limator and of the observing-telescope were replaced by non- 

 achromatic ones, whereby not only was there no loss of sharpness, 

 but, as was intended, by increasing the intensity of light, there 

 was a gain in clearness and distinctness. 



Sun. 



Parts of the screw. 



Deviations from the mean 



67-1 



-0-8 



69'4 



+ 1-5 



68-4 



+ 0-5 



67-9 



o-o 



66-6 



-1-3 



66-1 



-1-8 



68-2 



+ 0-3 



68-0 



+ 0-1 



69-6 



+ 1-7 



. . 67-9 ±0-3 





Mean 



Hence the distance of the two D lines would be determined 

 with a probable error of ^J 7T of its magnitude. From what has 

 been said above, an alteration of the distance between the source 

 of light and the spectroscope with a velocity of nineteen miles 

 in a second produces a reciprocal displacement of the lines of the 

 two spectra amounting to one-fifth of that distance — a magni- 

 tude, therefore, forty times that above found as the probable 

 error from the mean of nine readings. 



Hence if, in observing stellar spectra, a sufficient quantity of 

 light can be used, it may be definitely decided by tne way de- 



2B2 



