388 Preston—Fiadiation Phenomena in a Strong Magnetic Field. 
4678, distinctly separated to the naked eye on the photographic plate. The 
distance between the side lines of the triplet is sensibly the same for both these 
lines and measures, 0°56 mm. Taking the magnetic field at about 25,000 cgs. 
units, we may state therefore in round numbers that a field of 20,000 cgs. units 
will originate a difference of wave-length of one Angstrém unit between the side 
lines of the triplets 4680 of zine or 4678 of cadmium. In other words, the line 
4680, say, is converted by a magnetic field of 20,000 cgs. units into three lines 
of wave-lengths, 4680°5, 4680, and 4679°5 respectively. 
Using the formula given by Dr. Larmor* for the change of frequency produced 
by the magnetic field we have 
QreH 
Mv? 
Mm — Ny = ’ 
where m, and are the frequencies of the side lines of the triplet, viz. the fre- 
quencies of the two lines induced by the magnetic field, and v is the velocity of 
light. Hence for the difference of wave-length 6), we have 
On » QneH 
v7 M*? 
which, for the wave-length 4680 (of zinc) gives in round numbers 
e 
aa 10°? 
as the approximate value of the ratio of e to J/ in this case; or in electro-magnetic 
units, the ratio is about one to a million for this particular line. 
With such a resolution as this im the case of these lines, which belong to what 
we may term the normal triplet type, it resulted that, as one should expect, many 
of the lines which were previously unresolved became separated, and, further, 
the constitution of those which did not show as normal triplets was rendered much 
clearer. ‘Thus, if we refer to the photographs illustrating this Paper, we see that 
while the normal triplet 4680, say, consists of three distinctly separated lines, of 
which the middle one is the most intense, yet the neighbouring lines, such as 
4800 (the blue cadmium line observed by Zeeman and others) do not appear 
us triplets of this type. On the contrary, nearly all the light is sometimes 
concentrated in the two outside lines, leaving a weak middle, which exhibits the 
appearance, not of a single middle line, but of two weak middle lines, so that the 
resultant appearance of the modified line is a quartet, in which the two outside 
lines generally are much stronger than the two inside lines. Further, in some 
* Dr. J. Larmor, Phil. Mag. vol. xutv., p. 5038, Dec. 1897. 
