60 
Mr Purvis, The in fluence of a strong magnetic field 
to become weaker, sharper and narrower, and the strongest of 
them seemed to be divided into very diffuse triplets. On the other 
hand, there are four band-like lines 4649‘8, 4643‘6, 4620'9, and 
4444'7 which were widened but were too weak to analyse. 
There were numerous other lines which appeared to be divided 
or widened, but they were not sufficiently strong to measure ; 
whilst the lines 4295 - 94, 334053, and 3073 03 did not appear to 
be either widened or divided ; they became a little sharper and 
weaker. 
The following lines were divided into two. The two con- 
stituents were generally very weak, and their appearance was of 
two lines the outside edges of which were more diffuse than the 
inside. On analysis, some of them divided into quadruplets, but 
they were too weak to measure exactly. In the other instances, 
also, they were probably made up of several constituents. The 
measurements were taken between the centres of the aggregates. 
It will be noticed that there are some of these doublets which 
have the same value of dX/X 2 , and the general appearance of the 
divided lines is also very similar. 
X 
dX/X 2 
4779-10 
+ 1-18 
0 
-1-18 
| The outside edges were more diffuse than the inside edges. 
4590-20 
+ 1-15 
0 
-1-15 
| do. 
4529-73 
+ 1-29 
0 
-1-29 
| do. 
4315-70 
+ 1-74 
0 
-1-74 
| The constituents were equally sharp in both edges. 
4298-88 
+ 0-72 
0 
-0-72 
| The outer edges more diffuse than the inside edges. 
4289-25 
+ 1-38 
0 
-1-38 
| The constituents were equally sharp in both edges. 
428G-20 
+ 1-32 
0 
-1-32 
| The outer edges more diffuse than the inside edges. 
4012-58 
+ 0-81 
0 
-0-81 
| Both constituents very diffuse. 
