I. INTRODUCTION:The accurate determination of the radium content of ordinary matter has many applications in cosmology and geophysics, but is a difficult analysis to carry out. Two problems are involved, first the removal of radon from the material being tested, then the accurate measurement of this radon by means of its alpha ray ionization. The radon removed and measured is about 10-18 gm.
II. REMOVAL OF RADON FROM MATTER:The principles used heretofore for removing radon from liquids are modified somewhat, and the apparatus given a simplified and entirely new form. The estimated uncertainty of measurement is now less than 1%.
A new principle for removing radon from rocks, involving an entirely different and much simpler type of apparatus is presented. Powdered rock specimens are boiled without flux by direct heating to 1800°C in a vacuum, graphite resistance furnace. The gases are led to the ionization chamber through a simple electrostatic ion-trap which removes all the ionized molecules coming from the boiling granite. All moisture absorbents and reservoirs are eliminated from the gas line, a small insulation drying device in the ionization chamber protecting the amber insulation from deleterious vapors.
A systematic study of the percentage of the total radon removed as a function of (a) temperature, (b) time of heating, and (c) fineness of sample grains, showed that temperature is the only important factor and that 1800° effects complete radon removal. The duration of heating is 2 1/2 min., which is to be compared with one hour for the quickest of former methods.
III. MEASUREMENT OF SMALL QUANTITIES OF RADON:In the measurement of small quantities of radon, emphasis is laid on accurate knowledge of the background ionization due to cosmic and local radiation. The statistical variations in the background are shown to impose the natural observation limit in these measurements, and twenty possible sources of background are discussed with a view to minimizing these variations.
A new bifilar, vacuum, removable-head electrometer with multiple ionization chambers is described. The ionization chambers contain a small insulation drying room, thus permitting the use of moist gases.
The absolute or radium calibrations of the electrometer-ionization-chamber system by two radium standards and by the empirical formula of Duane and Laborde are critically discussed.
A new theoretical formula for the wall effect with alpha particles and for the radium calibration of emanation electrometers, in terms of their easily measured physical properties, is presented and shown to agree with the experimental calibrations involving radium standards.
The limit of precision of the bifilar electrometer is shown, and a new apparatus described which will extend the measurements to the natural observation limit. The nub of this apparatus is a sensitive string electrometer which measures only the difference in ionization in two identical ionization chambers both of which contain the same background gases, but only one of which contains the radon or thoron to be measured.
IV. DETERMINATION OF THORIUM:It is shown that the thorium as well as the radium content of matter can be determined in a single run. Like radon, thoron is removed from rocks by direct heating, the threshold temperature being 850°C for biotite granite. The determination of the temperature of complete removal and the absolute calibration of the electrometer in terms of thorium await the completion of an all-metal gas circulating pump, the design of which is briefly described.
V. BY-PRODUCTS OF THE INVESTIGATION:Three appendices discuss the construction and characteristics of three other furnaces developed in the course of the investigation. The vertical, split-tube, graphite resistance furnace described is particularly suited to a variety of chemical and mineralogical investigations. It maintains 2000°C on only 7 kilowatts of power input.
The first artificial production of cristobalite is tentatively asserted; micro-chemical confirmation analyses being in progress.
