aE 
SECTIONAL TRANSACTIONS.—A. 333 
The measurements thus have given practically a’—a”=az’—a;” not only for the 
monatomic helium but also for the diatomic hydrogen, and it may be inferred that 
a for this latter gas has a value for the translational energy which does not appre- 
ciably differ from the value which a has for the internal energy. 
Prof. N. Bour.—On Atomic Stability. 
The characteristic stability of atoms presents us, as well known, with a funda- 
mental limitation in the applicability of the concepts on which the so-called classical 
physical theories rest. An analysis of this limitation will be attempted in connection 
with a discussion of the difficulties disclosed by the recent development of the theory 
of the electron, especially as regards the problems of nuclear constitution. 
Discussion on The Unit of Atomic Weight (opened by Dr. F. W. Aston, 
F.R.S.). 
Dr. F. W. Aston, F.R.S. 
The idea conveyed by the words ‘ Atomic Weight’ is one of the oldest and most 
fundamental in chemistry. It is simply the expression of the fact that elements 
combine with each other in fixed and definite proportions. Thanks to the outstanding 
accuracy of the balance among early scientific instruments of precision, some of these 
ratios, more correctly called ‘ Combining Weights,’ could be fixed to about 1 per cent. 
over a century ago. Fifty years’ work pushed the accuracy some ten times as far, and 
as it advanced yet further and no serious indication of variations appeared it is not 
surprising that atomic weights were described as natural constants and generally 
regarded as the weights of the atoms themselves. 
Apart from mere theoretical speculation this idea was not seriously shaken till, 
in 1910, advances in radioactivity and positive ray analysis showed it to be on an 
insecure foundation. When, some ten years later, the mass-spectrograph had shown 
it to be quite untenable, chemists were forced to make an important decision in 
nomenclature. Should the terms ‘ Element ’ and ‘ Atomic Weight © be made to apply 
to the atoms themselves; in other words, should chlorine now be said to consist of 
two elements of atomic weights 35 and 37, or should the old meaning of the words be 
retained ? As a member of the Internationa] Committee I was strongly in favour of 
the latter. This procedure was ultimately agreed to and has, I think, been justified. 
Since that time much painstaking research has been undertaken, on the one hand 
to find if the atomic weights of complex elements varied with their origin, and on the 
other to achieve separation of isotopes artificially. The results have proved con- 
elusively that for all practical purposes the variation in nature is negligible, and that 
although partial separation can be achieved by extremely laborious operations the 
quantitative results are insignificant. 
The present unit of atomic weight O=16, chosen upon chemical grounds, gives a 
scale upon which many of the most frequently occurring elements are sufficiently 
close to whole numbers for all ordinary calculations of analysis. From the purely 
chemical point of view the recent discovery that oxygen has isotopes is clearly of 
little importance. The practical substandards of atomic weight determination, 
chlorine, bromine and silver have long been known to be complex to a much more 
serious degree. As, in all probability, most chemists will be dealing for an indefinite 
time to come with the present mixed but constant quality of complex elements, it 
seems needless to alter a standard which has figured in chemical literature for so 
long a time. 
The point of view of the physicist is an entirely different one. He is interested 
in the weights of the atoms themselves. Exact determination of the relative densities 
of gases is a step in this direction, but this involves purification, which is essentially a 
chemical operation. As a development of pure physics the subject is quite modern, 
for it is only during the last twenty years that any comparison of the weights of 
individual atoms has been possible. The original parabola method of Sir J. J. 
Thomson was, by 1912, made capable of distinguishing atoms differing by 10 per cent., 
and of comparing their weights with an accuracy of 1 percent. This was sufficient to 
suggest, though not to prove, that neon had isotopes 20 and 22. The first mass- 
spectrograph (1919) had a resolving power of 1 in 130 and an acouracy of about 1 in 
