75 



TABLE I {Continued) 



said arc-lines might possibly have to be attiibuted to a less good 

 observational power in the extreme red and violet, appears therefore 

 unfounded. 



The importance of the ground spectra made it desirable to perform 

 new measurements. The best method to obtain the first spark 

 spectrum of potassium in great purity and completeness we found 

 to be exciting the luminosity of very diluted incandescent potassium 

 vapour under the influence of very rapidly varying electrical forces. 



When our investigation was in progress, there appeared a publi- 

 cation by Mc Lennan ') on the spectrum of ionized potassium. 

 His tables present a close resemblance to ours, but in his Table T 

 Mc Lennan gives the lines which he has observed besides those of 

 ScHiLLiNGER. Heuce he also finds the arc-lines, which we succeeded 

 in eliminating. 



Besides both in Ids and in Schim.inger's observations a few important 

 lines are wanting. Important because they have been serviceable in 

 the search for the regularities to be mentioned presently. By the 

 aid of Table I it is possible to compare the measurements of our 

 second (Dj with those of the other observers, besides the data in the 

 column "remarks" show which P,Q, etc. could only be determined 

 by the new lines. At the same time it is at once clear which of the 

 lines are arc-lines. We estimate the accuracy of the measurements 



1) Mc Lennan, Proc. R. S. Vol. 100. 182. 1921. 



