STANDARDS OF MASS No attempt has so far been made to define a unit of mass by means of any natural standard, though a standard of this kind, e.g., a definite multiple, say, of the atomic mass of helium, would not be inconceivable. Prior to the discovery of the radioactive elements mass was regarded as the essentially constant attribute of matter, and there was no reason to anticipate any change in a material standard of mass except by actual damage due to abra sion, oxidation, hygroscopic absorption or other similar causes. And there still remains a reasonable choice of materials, which, given due care in preservation and handling, may be expected with considerable confidence to exhibit constancy of mass.
predecessors, igno rant of radioactive processes, were fortunate in the choice of platinum and platinum-iridium ( o% iridium) as the materials of construction for the ultimate reference standards of mass, the Im perial Standard Pound, and the International Prototype Kilo gramme, respectively. The degree of consistency (within i part in
with which recomparisons of various national copies of the kilogramme, made after the lapse of many years, have in general repeated the original determinations, speaks convinc ingly, not only as to the suitability of the standards themselves, but as to the perfection of the balances used in the comparisons. The relation between the two units, according to the best ascer tained determination, is i kg. = 2.2046223 pounds. This value has received legal sanction in Great Britain.
Another material presenting a high degree of constancy of mass is crystal quartz. This, however, has the dis advantage of having a comparatively low density. It also suffers from hygroscopicity, and requires careful handling to ensure the elimination of any effects due to surface electrical charges caused by cleaning.
In the comparison of two standard masses in air, allowance must be made for the upward buoyant effect due to the volumes of air which they respectively displace. The less the density of the mass, the greater will be the buoyancy correction. The accuracy attained in the intercomparison of a series of platinum-iridium standards is no doubt attributable to a considerable degree to the fact that they all have comparatively high, and very closely equal, densities, so that the net buoyancy corrections are very small, and a comparatively rough determina tion of the air density consequently suffices to give the correction with negligible error. In comparing a number of masses differing
appreciably in density, e.g., platinum, quartz and brass, the ac curate determination of the buoyancy correction presents much greater difficulty, and several attempts have been made to over come it by actually conducting the weighing in vacuo. This in volves enclosing the whole balance in an air-tight case, and manip ulating the weights entirely by mechanical means from outside, without opening the case. Leakage at the glands where the operat ing spindles enter the case has, however, so far proved an almost insuperable obstacle to successful weighing in vacuo. Everyday weighings for commercial purposes are of course necessarily con ducted in air, but the differences in buoyancy between the weights used, and the goods weighed, are negligible for this purpose. It is necessary, however, to provide a basis for the periodical reverifica tion by inspectors of weights and measures of traders' weights, which may be of iron, brass or other materials. For this purpose a "commercial" standard is employed. This standard is of brass (of density 8.143) adjusted to agree in vacuo with the Imperial Standard Pound of platinum. Inspectors' standards are also of brass, and all verifications of these standards, and thus indi rectly of traders' weights, are made by comparison, in air, with the commercial brass pound.
Even when weighings are not conducted in vacua the construction and manipulation of a balance for the accurate comparison of primary standards of mass are distinctly elaborate. It is necessary for the greatest care to be taken to preserve constancy of temperature, in order to maintain a steady zero reading of the balance. For this reason the room containing the balance must be thermostatically controlled, and the observer either works entirely from outside the room, or, if he enters it, must remain at a distance from the balance, all the manipu lation of the weights being effected from outside the balance case by mechanical control operated by means of long rods, and the movement of the balance beam being observed either through a telescope, or by the movement across a scale of a spot of light reflected by a small mirror attached to the beam.