7 6 



Let us suppose that we are dealing with the 

 spectrum of a star. Side by side on the photo- 

 graphic plate we have the stellar spectrum and 

 that of a spark between poles of iron or titanium. 

 By measuring these lines with a suitable instru- 

 ment, we obtain their positions within a few thou- 

 sandths of a millimeter. When the positions of 

 hundreds or, as in some cases, thousands of lines 

 must be determined with the highest possible pre- 

 cision, it is evident that the measurer must be both 

 painstaking and persistent. 



Next comes the work <>t computation. The 

 measured distances between the standard (spark) 

 lines and the corresponding lines of the star must 

 first be converted from fractions of a millimeter 

 into tractions of an angstrom (the international 

 unit of wave-length in spectroscopic work). When 

 this has been done, a short computation gives the 

 velocity of the star's motion in the direction of the 

 observer. After computing and deducting the 

 velocity of the observer's motion (due to the diurnal 

 rotation of the earth and its revolution about the 

 sun), the star's motion with respect to the sun is 

 determined. 



From this point one may go on, according to 

 the requirements of the work in hand, to identify 

 all of the star's lines with those of known chemical 

 elements or compounds; to account for minute 

 displacements of the lines from their normal places, 

 as the result of pressure or electrical excitation or 

 some other condition in the star's atmosphere; 



