8 B. E. Moore 



of the zero position. One could only answer this possibility by 

 photographing with very strong fields. It is not implausible, 

 because there are cases found in thorium where the number of 

 components on the two sides of the zero position do not balance. 

 One component of one line which is unseparated for the weaker 

 field strength is double for the 20 per cent stronger field. That it 

 is such a splitting is certain because the mean of the two separa- 

 tions is symmetrical with the separation of the opposite component. 



In one small region of the spectrum (Table XIX) I found it 

 necessary to make accurate measurements of wave-length, from 

 line to line, to make sure of identification. I have recorded these 

 lines as measured, although Exner's and Hascek's tables are more 

 accurate than my measurements. However, lines are often 

 shifted by small amounts, and these lines were peculiar in other 

 respects (see discussion, Table XIX). Hence there might be 

 cause for the shift, and I have, therefore, not hesitated to record 

 them as read. There can, in general, be no question as to the 

 identity of these lines, unless it be for lines not given in Exner's 

 and Hascek's tables. There were many of the latter, and it may 

 be an error to call them all thorium. Many of them possess pe- 

 culiarities in common with lines which are certainly thorium in 

 the list. There was no special effort made to find impurities, but 

 lines due to calcium, iron, strontium, and lead are present, and 

 none of these impurities behave peculiarly. Some of the strong 

 lines of Exner's and Hascek's tables are omitted because of their 

 weakness, but they are generally omitted because of the confusion 

 due to overlapping components of other lines, leaving the com- 

 ponents a blurred mass. 



The following abbreviations have been used : ^- and p- as al- 

 ready explained; r, red; b, blue; A, wave-length; d, or cliff., dif- 

 fuse; 0, order; 0, overlap; i, intensity; a, normal separation; A, 

 approximate separation ; AA/A 2 change in wave-frequency per 

 cm., and therefore a positive value, means a displacement toward 

 the blue end of the spectrum. The normal separation corresponds 

 to a value of 1.105 for a field intensity of 24450 lines to the square 

 cm., used in my previous contribution and checked here from the 

 separations of the calcium lines. If the strength of field is in 



96 



