1908-9.] Dissymmetrical Separations in the Zeeman Effect. 79 
rarely exceed (H)05 mm. The accuracy depends, of course, greatly on the 
character of the lines, their sharpness, intensity, separation, etc. The 
measurements have been made sometimes in the first, sometimes in the 
second, and at other times in the third order, according as the conditions 
were most favourable to accurate measuring. 
a and n denote respectively abnormal and normal dissymmetry. One 
can see that in most cases the intensities of the two outer components of a 
triplet are the same. In the above table are the most striking examples of 
dissymmetry in the spectrum of this element, but certainly there must be 
many more which are not so apparent to the eye. Only a few examples 
with more than three components are given, as it is desirable to limit the 
consideration to the simplest cases. 
It will be at once apparent from the table that concurrently with the 
change in the intensities there is a change in the type of dissymmetry 
found. As the middle component passes through its minimum value there 
is a change from the normal dissymmetry to the abnormal, and this change 
persists till the intensities pass through another critical point where the 
intensity of the centre component is twice as strong as that of the outer 
components. This would point to the rotation of the plane of polarisation 
due to the quartz being accompanied by a change in the type of dissym- 
metry. Against this it may be argued that there are lines between those 
cited whose planes of vibration would be so situated as to give dissymmetry, 
but that these lines have their components symmetrically placed. Again, 
there are exceptions to the rule that dissymmetry and plane of vibration of 
the components change correspondingly. For instance, A = 2606'50 shows 
abnormal dissymmetry ; and as the others in the neighbourhood show normal 
dissymmetry, this should also have been a normal type. These are points 
which at present I cannot explain, but the evidence for the connection 
between the dissymmetry found with the apparatus used and the rotation 
of the plane of polarisation is too strong to be overthrown by these excep- 
tions. There are also deviations from the rule established by the theory 
that the outer components have the same intensity. The lines A = 3373 - 88 
and 3429T5 show this, but the shadows on the violet side indicate that 
there is some other external cause to bring about this change. 
To show that dissymmetry could really be due to the rotation of the 
plane of polarisation, another test was made. The very bright line in 
molybdenum, A = 5533'26, was observed. By moving the quartz lens the 
beam of light was made to pass through different thicknesses of the quartz. 
This had the effect of altering the rotation, and hence also the intensities of 
the components. At the same time as the intensities altered from the outer 
