Jack — Magnetic llesolution of the S'jiectrum Lines of Niohium. 49 



but still an examination of the separations in the light of this rule may serve 

 to show whatever little value there is, and accordingly this is done. 



X = 2992-10 



A = 3484-25 



(0 = 1-35) w.= l jw =l-16")m=l 

 <o'=0 Jm=1, 21 w'= 0-46) 71=0,1 



X = 4249-61 

 w'=l-OlUi,= l 



X = 4331-56 

 w=0-64|m=l 

 a)'=l-56) 91=1 



; X, = 4345-48 

 la>=0-96|m = l 



lw'=l-73) n=l 



For the first four lines the agreement is good, but, considering the nature of 

 the test, it is not good in the case of the fifth line. 



There are nine lines irith five eomponents. 



Table IV. 



A = 2846-40 



A = -isss-gs 



A = 2931-.59 



i 



AA ! 



i 



AA 

 A- 



A 



i 



AA ! 



A» 1 ^ 



2 

 8 

 4 



+ 1-62 s ! 1 X 1-61 /^) 

 ± 1-59 p 1 

 3 



4 



6 

 8 



+ 1-12 p I 1 



p ; 



i 



3 

 3 



1 



i 1-85 s 



± 0-93 p 



p 



-«(y) 



1 







A = 2977-79 



A - 3014-56 



A = 8094-30 



i 



-" 



. ! AA 



1 



A 



i 



-^ i A 



2 



2 

 2 



± 1-42 s 4 

 p 



1 

 1 



± 2-43 s 2 X 1-22 (^) 

 ± 1-22 p 1 

 p 



4 

 25 

 50 



± 2-92 s 



± 1-42 s 



p 



2x1-447(1) 



1 







A = 3225-63 



A = 4137-27 



A = 4308-27 



. 1 AA 



' 1 -iJ 



A 



i 



""^ A 

 A' ^ 



■ 1 '^'^1 i 



13 

 13 

 50 



± 1-12 s 



± 0-37 s 



p 



3x.372g) 



1 

 



4 



7 

 7 



± 2-36 s 

 ± 1-21 p 



a 



1 

 



1 

 1 

 1 



+ 1-7.5 s 



± 0-51 p 



s 



7x-25lg) 



2 

 



Again, there are no lines having the same separation ; and therefore no 

 two of these can belong to the same series. For most of those lines there is 

 an obvious simple I'elationship between the amounts of the separations ; and 

 it is further satisfactory to find that these intervals are represented by 



R.I. A. PROC, vol.. XXX., SECT. A. ( 8J 



