TRANSACTIONS OF SEOTION A. 295 
and prevalence of a certain type is determined in a somewhat similar way— 
that some one type has been preferred. : 
Any theory of atomic structure to be satisfactory to chemists must take 
fully into account the peculiar valency relationships that are manifest among 
the elements, as the system of ‘structural’ formule now in vogue is based 
solely upon these. The system is admittedly one of extraordinary perfection 
and remarkably simple. In the case of organic compounds, the rules laid down 
have been found to be applicable and to suffice in so many thousands upon 
thousands of cases that it is impossible to doubt their general correctness; at 
most it will be necessary eventually to translate them directly into some new 
language. It should be pointed out, however, that so-called structural formule 
are to be regarded as condensed symbolic expressions indicative of the general 
behaviour of the compounds represented in terms of certain well-understood 
conventions, rather than as actual representations of structure. For example, 
it is customary to represent benzene by a regular hexagon, a symbol which is a 
complete expression of the chemical behaviour of the hydrocarbon. But the six 
carbon atoms are not to be thought of as arranged in a plane and in a ring in 
the manner depicted by the symbol; such an arrangement is impossible if the 
affinities of the carbon atom act tetrahedrally. The structure of benzene is 
rather to be represented by a model in which six carbon atoms (represented by 
six large spheres) are arranged three and three, in two superposed layers, union 
taking place between an atom in one plane with a contiguous atom in the plane 
above, which in turn is united to that in the plane below—so that the atoms 
are connected in zigzag fashion; and the six hydrogen atoms are to be thought 
of as severally united to the six carbon atoms in such manner that the hydrogen 
atom is always in a plane different from that which contains the carbon atom 
with which it is connected. If the ‘atoms’ in such a model are squeezed down 
into one plane, the projection is practically identical with the ordinary ‘ centric’ 
symbol of benzene. The arrangement referred to is shown in the accompanying 
figures (see Plate). 
The fundamental assumption made by chemists, upon which their system 
of structural formule is based, is that the hydrogen atom has unit valency— 
that it is incapable of acting as a linking element. The whole of the evidence 
available appears to be in favour of this view. The contention advanced 
recently by Sir J. J. Thomson, that hydrogen may occur as a triatomic molecule, 
Hs, is therefore unacceptable; until the existence of such a molecule has 
been proved up to the hilt it will be impossible for chemists to admit its 
existence. The artifice by which Sir Joseph Thomson has sought to reconcile 
his interpretations with those of chemists practically involves the representation 
of hydrogen as a dyad; if this conclusion were accepted it would be necessary 
to double the valency of all other elements, a step which cannot be justified on 
chemical evidence. It is in cases such as these that a better understanding 
between chemists and physicists is required. 
The variation of valency is probably the most perplexing phenomenon in 
chemistry. It is doubtful if any element have a higher ‘true’ or fundamental 
valency than carbon; the view sometimes put forward that certain elements may 
function even as octads is based on evidence which in no way justifies such 
an assumption. Not only will it be necessary to account for the variation in 
valency from element to element but also for the fluctuations observed 
especially in the case of the non-metallic elements. The variation seems to be 
determined by some reciprocal relationship between the interacting elements, 
valency apparently being a dependent variable rather than an absolute pro- 
perty; thus, to quote examples, whilst the hydrocarbon, CHz, is non-existent 
and cannot exist per se, the corresponding oxide, carbonic oxide, CO, is not 
only stable but relatively inert, combining with other substances only under 
special conditions; and the corresponding sulphur compound is so active that 
it cannot exist independently, but at once undergoes polymerisation with ex- 
plosive violence. Yet sulphuretted hydrogen occurs as a gas of simple mole- 
cular composition, whilst water, being a liquid of relatively high boiling-point, 
is presumably of considerable molecular complexity, so that it must be supposed 
that the fundamental molecule OH, is a highly active material. No theory of 
