Table of contents

The height distribution of the atmosphere depends on the temperature and the mean molecular weight of the air; recent progress in our knowledge of these two quantities will be reviewed, the main sources of information being outlined. The motions (winds and tides) in the upper atmosphere will similarly be considered.

The main gross uncertainty regarding composition at high levels relates to the proportion of atomic nitrogen present. Besides the major constituents that chiefly determine the mean molecular weight, there are important rare constituents, such as ozone, sodium and hydroxyl, which reveal their presence by their absorption or emission spectra; and there are others whose presence can be inferred from theirs.

The composition of the atmosphere changes by escape from the top (hydrogen and helium) and, by addition, from below (e.g. helium and carbon dioxide) and, as shown by the auroral spectrum, from above (hydrogen); cosmic rays also both add to the atmosphere and produce changes of composition by nuclear reactions.

Near the magnetic equator there are important ionospheric phenomena not yet fully explored and explaned. Among these one of the most interesting is the abnormal intensification of electric current flow, specially notable over Huancayo in Peru.

This paper contains a general discussion of electron drift in gases according to the method of free paths in which a fallacy contained in some earlier treatments is avoided. A general formula is derived for drift velocity in an electric field which is an arbitrary function of the time, and from it formulae for drift velocities in the following special cases of importance are deduced. drift in a constant electric field, in an alternating sinusoidal field, in a constant electric field with an applied magnetic field, in an alternating electric field with an applied magnetic field. Formulae for the electrical conductivity of an electron-laden gas follow from those for the drift velocities Applications of the formulae to ionospheric and other studies are mentioned.

The refractive indices of air and its principal constituents have been measured at a frequency of 24,000 Mc/s. with a precision comparable with that obtained in the optical range. The method is based on the measurement of the resonant frequency of a cavity resonator first when it is filled with the gas and then when it is evacuated. The source of oscillations used is a Pound-stabilized velocity-modulated oscillator, and its frequency is measured by reference to a high harmonic of a quartz standard, with a precision of 1 part in 10⁸. The cavity is provided with a tuning plunger, which changes the resonant frequency through a range of about 10 Mc/s, and is calibrated to an accuracy of within 1 kc/s.

Most of the measurements were made by the frequency-change method and the plunger was used in a narrow region only for a precise setting to resonance. It is possible by a larger movement of the plunger to compensate for the whole frequency change due to the removal of the gas and thus to work at a fixed frequency.

The following results were obtained for (n - 1)10⁶ where n is the refractive index at 0°C., 760 mm Hg: dry CO₂-free air, 288.15 ± 0.1; nitrogen, 294.1 ± 0 1; oxygen, 266.4 ± 0.2; argon, 277.8 ± 0.2; carbon dioxide, 494 ± 1; and the value for water vapour at 20°C., 10 mm. Hg pressure was 60.7 ± 0.1

The dielectric constants can be calculated from the relationship μ = n², where μ is the magnetic permeability and the dielectric constant, the values of (μ - 1) 10⁶ being taken as 0.4 for air, 1.9 for oxygen and zero for the other gases.

Accurate formulae are given for obtaining the refractive index of moist air at different atmospheric conditions, and are reduced to the following simple formula which is applicable for normal atmospheric conditions:

(n_t,p - 1)10⁶ = (103.49/T)p₁ + (177.4/T)p₂ + (86.26/T)(1 + 5748/T)p₃,

where p₁, p₂, p₃ are the partial pressures of dry air, carbon dioxide and water vapour, t is the temperature in degrees c., and T = 273 + t is the absolute temperature. A value of 1.839 ± 0.002 × 10^-18 E S.U. was derived for the dipole moment of the water vapour molecule.

The measurement of the alpha-particle activity of a `thick' radioactive source containing the natural radioactive series is discussed critically.

Two methods of determining the thorium-uranium ratio are considered for use in ore or mineral assays. The energy discrimination method is simple in procedure and can take account of lack of secular equilibrium between members of a series in the radioactive sample. The beta-alpha coincidence method, making use of the difference in the disintegration constants of radium C' and thorium C', is more complicated.

Initial experimental verification of the theory has been obtained. The estimated statistical error in determining the thorium-uranium ratio by the energy discrimination method is ±5% in the region of Th/U = 1, with a 4% sample (U₃O₈ equivalent concentration) in 4 minutes counting. A similar accuracy is obtainable in the coincidence method with a counting period 10 times greater

An expression is derived for the flow field in a two-dimensional jet of ideal fluid impinging normally upon a plate situated at an arbitrary distance from a parallel sided orifice. Numerical data are given when the orifice is at an infinite distance and in three other cases. In each example the trajectories of solid particles carried by the fluid are calculated by a stepwise method, and the conditions determining whether a particle will strike the plate or not are found.

The fluctuations in the energy lost by fast electrons over a fixed small fraction of their range have been studied by measuring the pulses the electrons produce in passing through a proportional counter. The energy loss distribution calculated theoretically by Landau is approximately gaussian with a high-energy tail. The experimental distribution is in agreement with the theoretical on the high-energy side of the peak, but is wider than Landau predicts on the low-energy side; this discrepancy is greater in krypton than in argon. The experimental values of the most probable energy loss agree with the theoretical values.

The current flowing through a 50 c/s, electrodeless discharge in a short cylindrical glass vessel with plane electrodes at the ends filled with chlorine at 5-50 mm. Hg was investigated with an oscillograph. It consists essentially of between one and fifty or more distinct pulses per half cycle, the number increasing with the voltage. At sufficiently large voltage these pulses cover the greater part of a half cycle including the instant of zero voltage. By irradiating the discharge with light from an incandescent lamp the average pulse height decreases, radiation below 4,800 A. being strongly active. By irradiating only the central portion of the vessel the pulse height does not change Irradiating one wall reduces the height of the pulses in every second half cycle, namely those pulses whose electron avalanches start from that wall

Once the discharge starts, electric charges collect on the inner surface of the vessel. Each current pulse consists of a series of avalanches which develop between small areas of the inner glass walls. One such area may discharge within a half cycle in several steps. The time variation of the applied field, the wall-charge field and the field in the gas are discussed and a new mechanism is developed which appears to apply to discharges between solid dielectrics.

At room temperatures, chlorine molecules form an adsorbed layer, several molecules thick, on the glass walls. Irradiation produces photo-dissociation of Cl₂, the atoms being effective in capturing electrons in this layer where slow electrons and many-body collisions are likely. This reduction of the number of secondary electrons reduces the pulse height. Adsorbed molecular gas layers also explain observations at higher temperatures At higher voltages the interval between two successive pulses becomes so small that one pulse reduces the height of the following one. This is probably due to metastable Cl atoms and resonance radiation diffusing to the wall where they cause dissociation of molecules in the adsorbed layers, the chlorine atoms capturing electrons as before. It also explains why at high voltages the effect of irradiation becomes insignificant