264 
Proceedings of the Royal Society of Edinburgh. [Sess. 
effect would not be large enough to mask the effect of the transverse 
field. It seems to me, therefore, that attempts to explain the supposed 
increase of resistance in low transverse fields are quite uncalled for. 
What requires theoretical explanation is the decrease of resistance of 
both iron and nickel in transverse fields, and the increase of resistance 
in longitudinal fields. 
This conclusion receives further support that in the case of cobalt 
Grunmach (2) obtained only a decrease of resistance. The cobalt was not 
in the form of a thin wire, hut was a strip 02 mm. thick and 0*5 mm. broad 
coiled in a double flat spiral. With such a form there was less chance of 
error of adjustment. Consequently no increase of resistance was obtained 
in the lower fields. 
With the doubtful exception of tin in the lowest field, all the other 
metals experimented with by Grunmach showed increase of resistance in 
transverse fields (2). These metals were silver, cadmium, tantalum, 
platinum, tin, gold, palladium, zinc, copper, and lead. 
Through the kindness and by the help of Principal A. Crichton Mitchell, 
late of Travancore, I am able to add to these mercury. Professor Mitchell 
prepared a thin mercury column in a spiral glass tube of a convenient size 
to be inserted in the air-gap of the electromagnet which I used for 
establishing the transverse fields in the present experiments. Substituting 
the mercury spiral for the iron or steel ribbon in the arrangement described 
above, I measured the change of resistance in four different fields. The 
results are given in the following short table, in which the first row gives 
the values of the transverse -field in Gauss, and the second the correspond- 
ing changes of resistance per 10,000. 
Change of Resistance of Mercury in Transverse Magnetic Fields. 
Tranverse Field. 
2064 
3801 
5263 
6473 
cZR/R. 10 4 
+ 0T1 
+ 0-31 
+ 0-43 
+ 0-64 
The relation between these sets of numbers is not linear, nor does a 
parabolic law satisfy them very satisfactorily. Nevertheless, assuming the 
formula dR/R = Af 2 , where t is the transverse field, we find 
A = 1'7 x 10~ 12 , 
a result of the same order as for other non-magnetic metals. 
[Note added November 19, 19 — My attention has been drawn to a 
paper published in 1910 in the Nuovo Cimento (5), in which Dr G. Rossi 
