400 Mr. G. W. Walker on Stresses 
of ferric chloride, so that the liquid stood at a certain height 
in the capillary tube. The bulb was placed between the poles 
of a strong electromagnet, and when the magnet was excited 
the level of the liquid in the capillary tube fell. The bulb was 
in the form of a disk, the direction of thickness being parallel 
to the lines of force. 
Mr. Walker argues that his system of stress would produce 
a hydrostatic pressure throughout the liquid “ under which it 
must contract, and the bulb slightly expand; whereas the 
stresses of electric type would only give a very small force at 
the surface of the liquid in the capillary tube which would 
make the column rise.”’ 
It appears to me that Mr. Walker’s discussion takes no 
account of the surface forces due to stresses of electrical type 
which must exist at the surface separating the liquid in the 
bulb from the glass, and, moreover, involves the assumption 
that the expansion of the glass under an internal pressure is 
of less importance than the contraction of the liquid under 
the same pressure. 
It is easy to show that the stresses of electrical type give a 
force per unit area directed outwards on the inner surface 
exactly equal to the hydrostatic pressure in the liquid on 
Mr. Walker’s theory. 
For assuming that the glass is non-magnetic and that the 
field is uniform in the liquid and outside, let H, be the mag- 
netic force in the glass, H, the magnetic force in the liquid, 
then 
where = is the permeability of the liquid. Then the force 
0 
per unit area directed outwards is 
1 2 lad ») 
races pert H,’) 
that is 1 ae ~1) Hen (2 —1) Hy; 
Sim pe \ Mo S77 fo \Mo . 
and since © is very nearly unity this is practically 
Ko 
1 (p 2 
My contention is that the expansion of the glass vessel 
under this force (of electrical type) is more important than 
