470 SECTIONAL COMMUNICATIONS. 
Since the sheaths of atoms of atomic number less than about twenty-five 
never contain more than eight electrons, the covalence of these atoms cannot 
exceed four, but with heavier atoms larger values might be expected. Large 
covalences are improbable in most cases, since they imply large negative 
valences, which means that the number of electrons in the sheath must be much 
larger than the charge of the kernel. There is evidence, however, that large 
covalences sometimes exist, the compounds Fe(CO); and Ni(CO)4, for instance, 
being complete compounds in which the central atoms have the covalences 
10 and 8 respectively. Since the number of electrons in the sheath of iron 
is 8 and of nickel is 10, the complete sheath in each case containing 18 electrons, 
the negative valences are 10 and 8 respectively, or the same as the covalences 
needed to account for the above compounds. 
The structure of a few substances is not adequately accounted for by the 
principles described, among them being N,,CO,CN—, and NO. These may 
have the simple octet structure described in the speaker’s earlier publications. 
The prediction that lithium hydride would exhibit the properties of a salt 
has been fulfilled, this compound having proved to ionise, hydrogen being an 
anion. In double molecules such as H,Oz (in ice), H,F,, and such compounds 
as KHF,, it seems that the hydrogen nuclei, instead of forming duplets with 
electrons in the same atom, form duplets in which the two electrons are in 
different atoms. The hydrogen nucleus itself thus acts as a bond in such a 
case. 
The theory thus outlined has been successful in accounting for the chemical 
nature of most compounds. The quantitative aspects are under consideration, 
and it is intended to put Postulates 1 and 3 into a form which will permit 
of at least rough calculations of the relative stabilities of various substances as 
measured, for instance, by the heat of formation. 
Prof, A. SmirHetts exhibited models illustrating the constitution of atoms 
according to Langmuir’s theory. The arrangement of electrons in layers was 
shown, and also the grouping of atoms in molecules to which the arrangement 
leads. For instance, the unioa of carbon atoms by means of electrons leads 
to a tetrahedral structure, already known from chemical evidence to exist. One 
series of models illustrated the structure of sodium carbonate historically, from 
the original arrangement of Dalton to that of Langmuir. 
Prof. W. L. Bracc.—X-ray analysis makes it possible to study the state of 
molecular aggregation in a solid and to find the positions of the atoms relatively 
to one another. The structural formula of calcium carbonate, on the ordinary 
hypothesis of valency, is of some such form as 
O 
ape 
but an examination of the solid compound shows that three oxygen atoms are 
similarly related to every carbon atom. The crystal consists of two kinds of 
units, calcium atoms and COs groups, the latter having perfect threefold 
symmetry, and there is no distinction between single and double bonds. 
The interpretation of the structure of graphite is not quite so certain as that 
of calcite, but the evidence is very strongly in favour of a hexagonal arrangement 
in sheets. 
Each carbon atom is equally related to three neighbours in the same plane. 
They are separated by a much greater distance from the atoms in other planes, 
and the cleavage indicates that the forces between successive planes are slight. 
These hexagons might be written as benzene rings with valency bonds, 
but such a formula would indicate a lowered symmetry of the structure, which 
is not the case. 
The work of Kossel, Lewis, and Langmuir has shown how valency may be 
partly explained. There is a sharp distinction between two cases. In the first, 
two separate electron systems, each surrounding a nucleus, are held together by 
electrostatic forces. In the second, the electron envelopes of each nucleus must 
be supposed to interpenetrate in some way certain electrons forming part of 
both systems. 
This view is supported by the evidence of crystal structure. In such a com- 
pound as solid potassium chloride the crystal consists of an alternate arrange- 
| 
. 
; 
. 
