274 
MAJOR A. E. OXLEY ON THE INFLUENCE OF MOLECULAR 
suggested that the electron couples constituting the molecule were of two kinds : 
(1) those of the atoms themselves, the sum of which presumably determine the atomic 
weight; (2) those of valency which alone are sufficiently free to vibrate synchronously 
with light waves and hence are particularly concerned in the refraction and dispersion 
of light. 
The valency or boundary electrons are vibrating under the control of the nucleus, 
but are less firmly held in the system than those near the nucleus. These valency 
electrons which have periods corresponding with luminous vibrations are affected by 
an external magnetic field in accordance with the well-known Zeeman and 
diamagnetic effects. The highly constitutive nature of the magnetic susceptibility is 
consistent with this view that the origin of the magnetic property is partly located 
near the molecular boundary. The nuclear electrons in the free atom will determine 
symmetry of the molecule, and are directly responsible, by their magnetic effect, for 
the symmetry of the crystalline grouping. This latter will therefore he determined 
by the nucleus, which controls the nuclear and boundary electrons, in an indirect 
manner and the distribution of atomic nuclei in accordance with crystalline symmetry 
as disclosed by X-ray methods is apparent. 
W. H. and W. L. Bragg have shown the difficulty, even in simple cases, of 
defining the molecular boundaries in a crystalline space lattice, although in some 
cases this is possible. But to determine by the X-ray method whether in any given 
crystal any atom has a special relation to a neighbouring atom would be practically 
impossible. The X-ray effects which they investigate are determined only by the 
nucleus or core of the atom and the o\iter electrons of the atom which contribute to 
its magnetic property, though they are controlled by the nucleus, are probably 
distorted by the influences of neighbouring magnetic elements. This distortion, which 
explains a large number of observed phenomena, defines the molecular boundary within 
the space lattice and determines a definite chemical molecule. These outer regions of 
the atom or molecule remain undetected by the X-ray experiments.^' 
* A. E. Oxley, ' Nature,’ No. 4, 1915. The core or electrostatic part of the atom is at a much greater 
distance from the atomic boundary than are the circular currents which give rise to the magnetic 
properties. As the intensity of the magnetic field due to a circular current varies inversely as the cube of 
the distance, and as in a crystalline structure two such circuits may approach so as almost to touch, each 
electron describing a small circle, the local magnetic forcive may be sufficiently large to account for the facts. 
The view that the cohesive force in crystalline media is of a magnetic nature was expressed in Part II. 
of this research, pp. 83-86. It was there stated that in a diamagnetic crystalline medium the molecules 
are held together by the local magnetic forces due to the revolving electrons. It is possible that each 
electron is completely bound to its own nucleus by a narrow tube of force, when the molecules would be 
capable of attracting or repelling one another electromagnetically according to their directions of rotation. 
The advantage of an electromagnetic cohesive force lies in the fact that by it we can readily see how 
similar molecules will cling together. Electrostatically such attraction implies an electron transfer 
which, we know, does not always take place. Reasons are given on p. 277 that the atomic forces which 
determine the structure of the molecule are in part at least of a magnetic nature. The advantage of 
