The Magnetic Separation of Lines 31 
The most of these lines are, as the footnotes indicate, prcebably 
of some other types, whose components have too small a separa- 
tion and are too diffuse to permit of an analvsis. The lines 
designated A are certainly not separated, and those undesignated 
in column B are probably not separated. The lines designated 
* in remarks are very curious and probably identical. The lines 
5730.0, 4240.39, 390c.71 would form a series, but the next term, 
about 3750, and the following terms are not present. 
VII. Osmium 
x i AX/X2 d Gawd ONY Se 
4420.63 25 +1.20 5 3882.03 7 +£1.65 
_ RS 25 10 p 3794.05 4 1.29 
4395.05 4 1.49 90.28 6° 1.83 
28.84 4 1.19 82.37 12 1.82 
11.15 6 ‘94 52.71 20) 1.63 
s 118 3598.26 4 [1.55] 
4294.17 8 AT 3561.04 8 1.35 
61.01 20) 1.95 59.96 4 1.58 
12.02 12 172 28.76 5 1.61 
4175.74 5 1.49 04.83 4 1.55 
73.42 10 144 3402.03 3 1.50 
35.95 15 1.97 3301.75 12 1.68 
12.17 10 1.65 3268.09 6 161 
4091.99 3 181 3962.44 4 1.37 
66.90 8 1.54 3156.38 4 1.62 
3977.38 6 154 3058.77 8 1.55 
63.79 8 1.59 2909.19 10 140 
Osmium was selected as a substance which, it was hoped, would 
give, with some types, repetitions througnout the spectrum, with- 
out having the confusion which arises from many lines of nearly 
the same separation. In this respect the substance was a disap- 
pointment. All the lines except one weak quadruplet are triplets, 
and the separation of these varies from .70 to 1.97. There are a 
number of pairs of similar separation, but not enough representa- 
tives for a series. The spectral lines were never strong. There- 
fore, a long exposure in the magnetic field was required. A char- 
acteristic of the components is their great sharpness. The un- 
aI 
