1878.] 
Suspended in Liquids. 
I 
I83 
cannot rest, as hitherto, upon experiments. We can only 
speculate that the action upon a minute irregular fragment 
will never be exactly equal all round. Differences of poten- 
tial will exist, owing, in the first place, to the non-homo- 
geneous nature of the particle. In accordance with all 
known laws of electro-dynamics, motion of the particle will 
result, which, by exposing new points of the particle to the 
action of fresh liquid, will alter the potentials, and lead to 
motion in a new direddion. In order that a particle shall 
rest motionless in a non-condudding fluid, it must be in 
exaddly equal chemical and eleddric relation to the fluid on 
all sides. That this should happen is almost infinitely im- 
probable. A condition of unstable equilibrium within limits 
is the result — a condition partially analogous to that of a 
ball suspended on a jet of water, which is always 
falling off one way or another, but is always brought back 
again. 
Having once begun to speculate it is easy to go a little 
farther, and to point out that there is probably a close con- 
nection between pedesis and the phenomena of osmose so 
carefully investigated by Graham. The conneddion is pro- 
bably that of action and reaction ; for if a liquid is capable 
of impelling a particle in a given direddion, the particle, if 
fixed, is capable of impelling the liquid in the opposite di- 
rection by an equal force, just as a steamboat, if not allowed 
to move forwards, will throw the water backwards. The 
earthenware jars used by Graham in many of his experi- 
ments consisted of silicious minerals, which would give 
pedesis in great perfection ; and Graham came to the con- 
clusion that whether the septum consisted of earthenware, 
animal membrane, or other substance, the motive force arose 
from chemical action upon the matter of the septum. The 
membrane was always corroded to some extent — a fact which 
quite accords with my explanation of pedesis. The solutions 
employed by Graham were also such as would admit of this 
view, containing usually about 1 per cent of the salt. It is 
obvious that if a pedetic particle be in contact with precisely 
the same kind of liquid all round, there would be no deter- 
minate direction of motion either of particle or of liquid ; but 
if one end of the particle be immersed in different liquid 
from that which surrounds the other end, a tendency to 
motion in a determinate direction will result, and this would 
be the condition of things in the porous earthenware jars of 
Graham. I do not pretend, however, to explain the exact 
modus operandi, which must be left to some one well versed 
in electrical science, but it is probable that the ultimate 
