750 REPORT—1885. 
(2) The dissociation hypothesis of Clausius and Williamson ; virtually 
accepted by Maxwell, modified by Quincke to explain migration, and 
shown by Kohlrausch to explain the facts of conductivity. 
(3) The electrostatic hypothesis of Helmhotitz. 
1. The Grotthus chain, with its postulated pairings and unpairings, is 
too familiar to need the smallest description. The only difference between 
the hypothesis as originally stated and the form in which I suppose 
Faraday to have held it, consists in whether the direct action of the 
electrodes be supposed to extend throughout the liquid, affecting and 
polarising every molecule in it,! or whether the electrodes’ direct action 
be considered as limited to a very minute molecular range, all inter- 
changes beyond this range being conducted on diffusion principles or by 
inter-actions among the fluid atoms themselves. 
The only objection that may plainly be urged against the theory is 
that it seems to require some small force able to effect the necessary 
initial decomposition, and it suggests that conductivity and tenacity of 
composition are related to one another in some opposing manner. Facts, 
however, fail to bear out any such idea; conductivity and chemical 
tenacity seem independent of one another; and, as has been just said 
under head ‘ polarisation,’ no finite force, however small, has ever been 
found necessary to decompose an electrolyte when really applied to it. 
In other words, no polarisation exists inside a homogeneous electrolyte ; 
there is no chemical cling of the atoms there, but only a frictional rub. 
Such a fact as this, if well established, renders necessary some form of 
dissociation hypothesis. A Grotthus chain of quite equidistant atoms 
might serve, or a momentary dissociation would be sufficient, but no 
hypothesis which involves a tearing asunder of molecules in the interior 
of a homogeneous electrolyte can be permitted. Herein lies the great 
distinction between electrolytes and dielectrics. 
2. The form of dissociation hypothesis suggested by Clausius and Wil- 
liamson is well known. It supposes that the vast majority of molecules 
in an electrolyte are quite insusceptible to the influence of electrodes, but 
that a few of them (the number being increased by complexity of com- 
position and by rise of temperature) are, by collision or otherwise, dis- 
sociated, and exist in the free atomic state, each atom with its appropriate 
charge. These alone feel the influence of the electrodes. According to 
some statements of this hypothesis the direct influence of the electrode 
is supposed to reach every dissociated atom in the fluid; according to 
another the direct electrode action extends only to those atoms which 
come within a minute range of its surface, everything else being managed 
by ordinary diffusion, 7.e., by the ordinary chance locomotions common 
to all atoms in a fluid. 
Individual atoms, though permitted to recombine as soon as they 
like, on this theory, are commonly thought of as existing in the disso- 
ciated state for a finite time. If there are chemical or other objections 
to such a view, it need not be held; all that the facts of electrolysis 
require is the most momentary dissolution of partnership, a temporary 
but quite perfect freedom, so that the feeblest possible influence may 
suffice to induce recombination in a definite direction and with some 
‘atom other than its former partner. Provided a sufficient supply of such 
! Grotthus in 1805 supposed electrodes to attract ions according to an elemental 
‘inverse square law. 
