14 B. E. Moore 



There is a characteristic of these triplets that is impressive. 

 One may choose a line with almost any magnitude of separation, 

 and not far from it find one or more lines which have practically 

 identical separation. The following seven lines lie in a small 

 bundle upon the plates. Four of these, and possibly six, are du- 

 plicates. These lines are 4383.69 A to 4374.96 A inclusive (see 

 table). There is another bundle between 4286.90 and 4283.25 

 with three lines alike. The photographic plates suggest many 

 such bundles or groups. Naturally, other types than triplets may 

 be present in these bundles. These at once suggest those recur- 

 ring groups in those substances in which series have been found 

 This fact, together with the numerous close companions having 

 like separations, would suggest that hope for relationship among 

 the thorium lines was not forlorn. 



Table XVIII is a list of unsymmetrical triplets not included in 

 Table XIX. Stronger exposure may reveal other components 

 for 3294.76. The line as recorded is uncommon rather than un- 

 symmetrical. The ^-component is in the zero position and the 

 /'-components separated. 



Table XIX contains a list of lines in a very small portion of 

 the spectrum. The spark spectra was obtained from thorium 

 chloride upon carbon electrodes. About 3885 A a carbon band be- 

 gins whose lines are troublesome for some distance. As usual, 

 I was omitting these lines of zero separation, but found that I 

 was omitting many lines from Exner's and Hascek's tables. Close 

 measurement of some of these unseparated lines showed that they 

 agreed so closely with the tabulated wave-lengths which I was 

 omitting that there could be no doubt as to their identity. Those 

 which were found separated were unsymmetrical. Unseparated 

 lines are supposed to arise from compounds. The carbon band 

 lines which are unseparated are attributed to cyanogen. A few 

 of the lines in this region Exner and Hascek mark cyanogen. 

 However, all but one of these show separation. Many not at- 

 tributed to cyanogen fail to show any separation whatsoever. 

 These may be due to compounds, but it is only conjectural. Their 

 behavior is generally very peculiar. In my study of zirconium, 

 I noted that most of the lines of zero separation looked like unre- 



102 



