June 30, 1923] 
- 
NATURE 
885 

he writes continuous lines between the symbols, 
and in 1868 puts the matter thus: “The structural 
formula of formic acid [V] indicates first, that the four 
carbon equivalents form one atom, the four oxygen 
equivalents two atoms, and the two hydrogen equiva- 
lents two atoms ; secondly, that these equivalents are 
united in pairs, thus—co, co, co, ch, ho; but it does not 
in any way indicate (and we do not know) what is the 
potential of each of these pairs—that is, how much 
energy would be required to separate the equivalents 
from each other. We know that this potential depends 
upon the structure, and we can to a certain extent 
trace the nature of this dependence, but as yet we 
cannot express the potential numerically, and _ till 
we can do that we do not fully know the constitu- 
tion.” 
On one hand, then, our “ bonds ” stand for mere 
units of valency ; on the other, they are an imperfect 
representation of forces. Were the representation of 
forces more complete, methyl alcohol would appear 
somewhat as shown in the figure below, the small letters 
representing numerical values. Even this formula, 
however, only gives the magnitude and not the real 
direction of the forces, and is besides static, not kinetic. 
We naturally shrink from complexity such as this: 
imagine the formula of sucrose on a similar basis. We 
Must content ourselves with something simpler, and 

Methyl Alcohol. 
yet the simple valency formula has for long been felt 
to be inadequate. Apart from the idea of definitely 
directed valencies which leads to stereochemical formule, 
the idea of representing partial valencies has been 
_ constantly in the foreground. We cannot properly 
split valencies in the old sense, but we can subdivide 
forces ad libitum. If the subdivision is carried too far, 
however, the formule may approach in complexity. 
formule with quantitatively measured forces, such as 
that indicated in the figure. 
What the chemist requires in his system of formula- 
tion is something, not which he can measure, but 
which he can count—counters, in short. Such numerical 
counters he possesses in valencies, in co-ordination 
numbers. He may be forced to consider the adoption 
of counters of different kinds, for the purpose of 
representing essentially different distributions of force ; 
but these counters, if they are to be of general practi- 
cal value, must neither be too numerous nor of too 
great variety. Partial valencies, augmented valencies, 
diminished valencies, virtual valencies, represented by 
lines of various sorts, thick, thin, curved, dotted, etc., 
all tend to complicate formule, which lose in obvious- 
NO. 2800, VOL. 111] 
ness what they gain. in definition. The humiliating 
fact must be admitted that the average man does not 
succeed in counting quickly and accurately by in- 
spection any larger number of “ bonds ” than he has 
fingers on one hand, unless they are appropriately 
grouped—witness the wrong valencies found, not 
merely in examination scripts, but even on the printed 
page. We ought, then, to set strict limits to the 
Splitting of “bonds” and the issue of fractional 
valency counters. Fortunately, the physicist has pro- 
vided us with a new counter, the electron, which has 
the great merit of being a physical reality, and, 
moreover, of being an undecomposable magnitude, so 
that there is no temptation and no possibility of 
dividing it further. This counter enables us practically 
to halve ordinary valencies, and so offers many ad- 
vantages. In the original form given by J. J. Thomson, 
lines in the formule are made to represent tubes of 
force passing from an electron of one atom to the 
positive core of another, and since in electrically neutral 
atoms, for each tube of force which passes out of an 
atom a return tube must come in, directly or indirectly, 
one valency line on the old scheme is represented by 
two tubes of force on the new. The new formulation 
enables us to write, for example, symmetrical formule 
for benzene and for the carboxy] group, if we so desire, 
thus : 
H 
i 
H=c% So=n O 
Mm lil = Si 
= sae \ 
H YWoG H Yo7% 
i 
H 
Or, again, we may represent the valency electrons 
directly in our formule, each atom being equipped 
with its doublet, sextet, or octet. Such formule, or 
modifications of them, are coming extensively into use 
when multiple bonds are in question, and there is no 
doubt that with the electron or electronic tube of force 
as counter we obtain a more adaptable and more 
flexible formulation of organic compounds than that 
afforded by the older valency formule, although only 
with some sacrifice of their simple character. 
It will be gathered from what I have said that my 
plea is for the utmost obviousness of our symbols and 
formule. Their content and connotation may be as 
rich as we please ; the symbols themselves should be 
of elementary simplicity. But, it may be asked, why 
should we seek to limit the investigator striving for 
chemical self-expression to four whole valencies for 
carbon if he wants a dozen partial valencies to facilitate 
his own thought or to convey his exact meaning ? 
And why should he hesitate to adorn his formule with 
arrows or positive and negative signs of polarity if he 
feels the necessity? The answer to such questions 
‘must be of a practical nature. The investigator for 
his own use may employ a symbolism as elaborate and 
as complex as he chooses, but if he wishes to secure 
the understanding and sympathy of others he must 
curb any spirit of exuberance. A system of chemical 
formulation to have general currency must not be too 
elaborate. Otherwise, although it effects economy of 
