on the spark spectra of titanium, chromium and manganese. 5 ( J 
X 
dX/X 2 
3058-20 
3056-86 
-i- 2-18 
0 
-2-21 
| The line widened into a broad band, which separated into 
J three very weak and diffuse constituents when analysed by 
( calcite ; they were too weak and diffuse to measure. 
3046-79 
+ 2-00 
0 
-2-02 
3029-84 
+ 1-20 
0 
-1-21 
3017-30 
2984-85 
+ 1-40 
0 
-1-39 
(The line separated into three very close constituents which 
were too diffuse to measure; and the calcite did not sepa- 
( rate them sufficiently well to measure. 
2956-26 
+ 1-88 
0 
- 1-88 
2948-40 
| 
2891-18 
l The constituents were all too weak and too diffuse to measure. 
2889 -00 
2884-22 
+ 1-60 
0 
-1-61 
2877-52 
+ 1-33 
0 
-1-35 
2862-47 
+ 1-38 
0 
-1-40 
2856-22 
Too weak and too diffuse to measure. 
Too narrow and diffuse to measure, and on separation by 
calcite it was too weak. 
2851-21 
2842-01 
+ 1-52 
0 
-1-54 
2832-24 
do. 
There were other lines which became triplets in the more re- 
frangible regions beyond 2832‘24, but they were all too weak and 
too diffuse to measure exactly. 
r i he spectrum of titanium has a number of lines more like 
diffuse bands, which, generally, are sharp on the more refrangible 
edges and diffuse and weak on the less refrangible. They appeared 
