30 Recent Chemical Researches . [January* 
or the number of positions, in which this force may be 
exhibited is the chemical valency of the atom. 
Now it is not necessary to suppose that the total force is 
exercised equally in each of the various directions ; it is 
possible that a greater amount may be exercised in direction i 
than m direction 2 : hence there may be a difference in the 
nature of a compound according to the direction in space 
of the lines joining the centres of the component atoms. 
The mutual aCtion of atoms is supposed by Michaelis to 
be for certain distances an attractive aCtion, but for very 
small distances a repulsive aCtion. Hence when two atoms 
come within the sphere of one another’s aCtion they ap= 
proach one another until the repulsive aCtion comes into 
play, when they are repelled until the repulsion gives place 
to attraction, and so on. But under certain circumstances* 
which may easily be imagined, the atoms are driven so far 
asunder that return within the sphere of one another’s 
aCtion is no longer possible ; the compound, therefore, is 
broken up. 
In this view chemical valency is constant for each element, 
inasmuch as the number of directions in which the chemical 
force may be exerted is a constant number ; but the force 
need not be equal in amount in each direction, nor need it 
be always expended in doing work in each direction when 
the element unites with others to form a compound sub- 
stance. Furthermore, the valency is independent of the 
total affinity of the element, so that when we speak of tetra- 
valent carbon and divalent oxygen we do not mean it to be 
inferred that the potential energy of the carbon atom is 
double that of the oxygen atom, but merely that the carbon 
atom is able to exercise its total energy in twice as many 
directions as the oxygen atom is able to do. 
The chemical valency of the elements may therefore form 
the basis of a system for their classification. In order to 
determine the valency of any element it is necessary to take 
into account the whole of its compounds so far as we are 
acquainted with them. But inasmuch as we cannot deter- 
mine with certainty the molecular weights of any compounds 
other than those which are gaseous at attainable tempera- 
tures, we must limit our attention to such compounds. 
Conclusions concerning valency, drawn from a consideration 
of solid or liquid compounds, are very misleading. 
The existence of so-called molecular compounds is liable 
to mislead in the determination of the valency of an element. 
No definition, which may be easily and exactly applied, of a 
molecular compound as distinguished from an atomic 
