Adeney — Photographs of Spark- Spectra. 25 



of the same metal have been employed as standards for calculating 

 the wave-lengths from the micrometer measurements. 



In the cases where the calculated wave-lengths have shown a 

 close agreement with those by Kayser, the latter observer's values 

 have been adopted. In those cases in which they have not agreed, 

 they have been confirmed or corrected by remeasuring the lines 

 with the same microscope and micrometer, but with the addition 

 of a finely divided glass-scale in the eye-piece, and a photograph 

 of a reseau belonging to the observatory of Cambridge University, 

 for which the author is indebted to Mr. Arthur B>. Hinks, m.a. 1 



With this arrangement, it has been possible to measure a 

 number of neighbouring lines relatively to one another with 

 considerable accuracy. 



In the appended list of wave-lengths, 2 Kayser's have been given 

 for convenience in indicating those lines which are common to 

 both forms of the spectrum and those which only appear to belong 

 to one or other form. 



In the author's photograph of the ruthenium spectrum, 1461 

 lines have been measured between the two extreme limits of wave- 

 length 2263 and 4560 ; and Kayser has given 1613 lines as occur- 

 ring in the arc-spectrum between the same limits of wave-length. 



About 800 lines are common to both forms of the spectrum. 

 Besides these, about 800 lines occur in Kayser's list, but appa- 

 rently not in the author's, and about 650 in the author's and not in 

 Kayser's. 



Displacements in some of the spark-lines, as indicated by 

 Haschek in his interesting paper entitled " Spectroscopic Studies," 

 may account for some of these differences; and such lines may 

 therefore be similar, although slightly differing in wave-length. 



Some of the wave-lengths in both lists may be affected by 

 unusually large errors, owing to diffuseness in character of the 

 lines, or to faintness or to some distortion in the films, or to some 

 other source of error, and may in reality be coincident. 



There can, however, be no doubt that a large number of the 

 lines, which appear peculiar to each form of the spectrum, are due 

 to the different modes of producing incandescence employed. 



1 See Mr. Hinks' paper on " The Cambridge Machine for Measuring Celestial 

 Photographs." Monthly Notices of R. A. S., May, 1900. 



2 Note the wave-lengths have been given to two places of decimals only. 



