28 JOURNAL OF THE 
est to a higher. The permanganate is of course. deoxidiz- 
ed. | 
It seems that chemical action may induce change when to 
an existing molecule a third substance is offered capable of 
combining with one or more of its constituent atoms, thus re- 
leasing the former equilibrium. Thus when sulphurous acid 
takes the oxygen of water setting hydrogen free the hydro- 
gen then takes one of the chlorine atoms held by the iron. 
FeCl,+H,SO,+H,O=H,SO,.2HC1+2FeCl,. 
But the presence of all three of these molecules is needed 
for the reaction to take place. So too, potassium perman- 
ganate is stable in the presence of sulphuric acid, unless the 
ferrous sulphate or some such molecules are present. When 
molecules of these three substances come together there is 
immediate rearrangement of molecules with change of equil- 
ibrium. Whether we are dealing here with a play of affinity 
which causes the tumbling down of certain molecules and 
building up of others, or whether it is a question of vibra- 
tory equilibrium between these molecules, cannot yet be told. 
The only certain thing seems to be that a molecule contain- 
ing bivalent iron and another containing septivalent man- 
ganese cannot exist in the presence of one another but must 
change, when possible, to trivalent iron and bivalent man- 
ganese. 
As meagre as our present knowledge is, it does not seem to 
be a very hopeful task to enter the maze of changes of val- 
ence through chemical reactions with a view to clearing up 
the ideas as to the nature of valence. 
Explanations Offered.—Victor Meyer and Riecke have sup- 
posed that a solution of the problem could be arrived at best 
by studying the phenomena of frictional electricity, contact 
electricity, pyro-electricity, and electrolytic conductivity. 
Most of those who have suggested hypotheses have based 
them upon a study of the carbon atom and its compounds and 
in particular its space relations. I have gathered together 
such of these hypotheses as have come to my notice. 
The first in point of time is the hypothesis of van’t Hoff’. 
1 Ansichten iiber die organische Chemie I. 3. 
