Index of content:
Volume 23, Issue 9, 01 September 1952
23(1952); http://dx.doi.org/10.1063/1.1702353View Description Hide Description
There are two types of crystal edges in the electron micrographs of oxide replicas of metals. The first type is characterized by the edges being more transparent than the facets in reprinted crystals. In the second type of reprinted crystal there were found the opaque edges of crystal. These characteristics of crystal edges were explained by the orientation of crystallites composing the replica films. It is proposed in the present study that the width of transparent edge of reprinted crystal might serve as a rough measure of microscopic magnification.
23(1952); http://dx.doi.org/10.1063/1.1702354View Description Hide Description
The relation between the crystal size of barium‐strontium oxide and its thermionic emission was studied with the same vacuum tube. The crystal size was measured by an x‐ray diffraction method. The measurements were made in vacuum by using thin mica windows as the entrance and exit ports for the x‐ray beam. The emission was measured with pulse in order to avoid gas from the anode. Below 1050°C the crystal grows slowly, but above it crystal growth is rapid. It seemed that impurity activation is dominant at low temperatures, but heat activation is dominant at high temperatures.
23(1952); http://dx.doi.org/10.1063/1.1702355View Description Hide Description
A new theory of turbulent diffusion is developed so as to describe the dynamics of particulate matter in fluid suspensions. In order that the statistical description be consistent with force equationdynamics of sedimentation, a new diffusionequation is derived from the postulates of a past‐future stochastic process.
A steady‐state suspension of particles in a turbulent stream is treated like an atomosphere. Dimensional arguments are used to retain the most relevant physical entities in a linear theory. It is shown that the bed load and mid‐stream suspension of a muddy stream cannot be considered separately. Some experiments are discussed in this connection. A qualitative theory of sand rippling is explained in terms of the instability of a flat sandy bottom under certain turbulence conditions. A singular perturbation method is employed to obtain an asymptotic expansion of the solution of the equilibrium equation.
Assumptions are discussed and criticized. The advantage of the theory is that all statistical coefficients are averages of familiar dynamical quantities.
23(1952); http://dx.doi.org/10.1063/1.1702356View Description Hide Description
The radiation from an infinite dielectric slab fed by a line source has been calculated from a branch‐cut integral. The modal efficiency of the device as a wave guide is defined as the ratio of the power propagated in a mode to the sum of the modal and radiated powers. For a range of the parameters such that only one mode can be propagated, the modal efficiency has been plotted numerically. Except for extreme values of the parameters, the efficiency runs around 80 percent to 90 percent.
The slab form is not one likely to be used in practice, but the efficiency of a cylinder would be expected to have about the same value.
23(1952); http://dx.doi.org/10.1063/1.1702357View Description Hide Description
Silver from the vapor phase was diffused into thin gold wires and foils at 940°C. Measurements were made of the resulting dimensional changes which occurred normal and parallel to the direction of diffusion. The linear percentage expansion normal to the diffusion direction was always found to be only slightly less than the corresponding expansion parallel to the diffusion stream. The results are best explained by the creation of new lattice sites at dislocations and support both the vacancy and interstitial mechanisms of diffusion in substitutional solutions.
23(1952); http://dx.doi.org/10.1063/1.1702358View Description Hide Description
A new device for metallic shadow‐casting by using the nozzle system is proposed, and an outline of the apparatus is given. The present device has a shutter, a nozzle, and a cover plate between specimen and filament in addition to the usual apparatus. The dimensions of the nozzle were chosen from the maximum of 3‐mm diam, 1.5‐mm diam, and 22 mm to the minimum 1.5‐mm diam, 0.7‐mm diam, and 38 mm as diameter at the entrance, diameter at the exit, and the length, respectively. It was found that this method was not only conspicuously effective in obtaining much more accurate shadows than that obtained by the usual method with the same shadowing material, but also brought into possible use metals, as shadowing materials, which could not be used for such a purpose by the usual method. Shadowing qualities of many kinds of evaporated metals are compared and a new rule for the order of aggregation of depositedmetallic thin films is also proposed.
23(1952); http://dx.doi.org/10.1063/1.1702359View Description Hide Description
The theory of Carlson and Heins and the extended theory of Lengyel has been verified experimentally for oblique angles of incidence. Both single surface and slab measurements were performed in the 3‐cm band. Single surface measurements were made by using tapered wooden wedges as absorbing materials to eliminate back surface reflections. The effect of plate thickness was studied for five different thicknesses. The phase and amplitude of the reflection coefficient of a metal‐plate slab were obtained for three angles of incidence, 10°, 25°, and 35°, and for values of 2a/λ ranging from 1.10 to 1.30.
23(1952); http://dx.doi.org/10.1063/1.1702360View Description Hide Description
Fundamentally, the method is closely related to one described by W. Weber in 1852 in his Elketrodynamische Maasbestimmungen and applied by him to the measurement of very small susceptibilities in cases in which it was desirable to exclude eddy currents from the specimens under examination. It is capable of being modified so as to measuremagnetic moments and magnetic susceptibilities, large and small, under a great variety of conditions. The method has the advantage, for low permeabilities, over that of permeameters, that it measures precisely the (small) difference between the permeability of the specimen and that of air, and not the ratio of the two. It determines the magnetic moment of the specimen and the susceptibilityk, from which the permeability μ may be derived from the relation μ=1+4πk.
23(1952); http://dx.doi.org/10.1063/1.1702361View Description Hide Description
The equation logη=A+B/(T+C) is obtained as a simple modification of Andrade's viscosity equation by treating the activation energy as a function of temperature, A, B, and C being free parameters. This is shown statistically to express adequately the temperature dependence of viscosity over substantial ranges of temperature for all the liquids examined, viz., n‐heptane; n‐nonane; n‐undecane; n‐tridecane; n‐heptadecane; benzene; viscous mineral oil; ethanol; 1,2‐dibromopropane; water; mercury; molten gallium;potassium;sodium;sodium nitrate; silver bromide; lead chloride; bromine; 98–100 percent sulfuric acid.
23(1952); http://dx.doi.org/10.1063/1.1702363View Description Hide Description
When the anode potential of a nonoscillating magnetron is lower than the cut‐off potential, the discrete electron‐to‐electron interaction has an effect on the space‐charge distribution which would be negligible if only resulting in fluctuations about a known steady state. However the states proposed by Brillouin, Slater, and others are not steady states in this sense. The electron interaction must, therefore, be expected to produce a steady drift away from such initial states. This paper discusses qualitatively the final steady state to be expected and its dependence on the ratio of anode voltage to cut‐off voltage. Because of the involved nature of the electron distribution function no attempt is made to calculate the space‐charge distribution explicitly.
23(1952); http://dx.doi.org/10.1063/1.1702364View Description Hide Description
This paper describes both a traveling‐wave method and a standing‐wave method for determining complex dielectric constants of liquids from measurements made at 10‐cm wavelength with a traveling probe immersed in a dielectric‐filled slotted line. An interferometric null technique was developed for improving the accuracy of the measurements, especially in highly absorptive media. In the traveling‐wave method, absolute measurements of both wavelength and absorption index can be made simultaneously; from these two quantities the complex dielectric constant is readily computed. In the standing‐wave method, the absorption index is determined graphically from the power‐standing‐wave ratio while the wavelength is obtained by a modified null technique. For low loss media, a simplified standing‐wave method is practical whereby the absorption index is obtained from measurements of the breadths of successive minima of the standing wave. Corrections for the finite resistance of the short circuit used in the standing‐wave method and for conductor losses along the slotted line are fully treated.
23(1952); http://dx.doi.org/10.1063/1.1702365View Description Hide Description
The paper deals with the propagation of unattenuated elastic waves in a cylindrical bore through an elastic material of infinite extent filled with a fluid. The phase and group velocity dispersion curves are plotted for axial symmetric waves in the coupled fluid‐solid system.
23(1952); http://dx.doi.org/10.1063/1.1702336View Description Hide Description
The effect of Co60 γ‐irradiation on the direct currentconductivity σ of polyethylene has been determined up to an intensity I of 4000 roentgens per hour at room temperature. The equationdescribes the data where σ00 is the conductivity without radiation and is the order of 10−18 (ohm cm)−1 and I 0 is 20 roentgen/hr. Measurements at liquid nitrogen temperatures give a factor of 20 decrease in the normal conductivity and a factor of 25 decrease in the radiation induced conductivity below their respective room temperature values.
The temperature data serves to suggest an ionic mechanism for electrical conduction in polyethylene as opposed to an electronic mechanism. In fact, the ion contributing most to the conduction in the polyethylene may well be the proton.
23(1952); http://dx.doi.org/10.1063/1.1702337View Description Hide Description
A new theory of metal transfer and wear is suggested in this paper. The essence of the theory is as follows: Metal transfer and wear take place at points of actual contact. The interfaces of the high spots that actually make contact are roughened as the result of plastic deformation when they carry normal load. The mechanical interlocking effect of these roughened interfaces is the primary cause of metal transfer and wear. Due to the mechanical interlocking effect of the roughened interface, and the strain‐hardening that accompanies plastic deformation, the application of a tangential force will break one of the pair of the contacting high spots a certain distance away from the interface rather than at the original interface. A secondary cause of metal transfer is the adhesion or the diffusion process which takes place during the temperature flash that occurs during breakage. If the adhesive force is very weak and the diffusion process is not rapid enough to cause the sheared‐off peak of the high spot to become a blob of transferred metal, the small piece of metal sheared from the high spot can leave as a loose wear particle. This proposed theory explains the welding of the sheared‐off peak to its opponent high spot as the consequence of friction, whereas, in the ``welding'' theory of friction,welding is considered as the cause of friction. Most metallic surfaces in ordinary atmosphere are always covered by a surfacefilm. The effect of surface contamination on metal transfer, wear, and the shear component of friction is discussed. Difficulties encountered in applying the ``welding'' theory to explain the friction of and metal transfer between contaminated surfaces where metallic adhesion is absent, are obvious. Experimental support of this new theory is given here. It includes as direct evidence (1) the roughening of the interface as the result of the plastic deformation, (2) the perfect matching at the roughened interface, which gives a strong mechanical interlocking effect, and (3) a definite region of severely strain‐hardened metal near the interface.
23(1952); http://dx.doi.org/10.1063/1.1702338View Description Hide Description
In current theories of the concentration dependence the solute molecules are treated as point centers of disturbance. The finite size that produces a shielding of two interacting particles in the presence of intervening ones reduces the interaction calculated in the point approximation and thus reduces the discrepancy between experiment and theory. An approximate treatment for a spherical solute based on the cage model, and hence more adequate at high concentrations, is presented. It operates with a spherical enclosure around a central particle. The position of this shell reflects the relative extent to which particles beyond nearest neighbors can hydrodynamically interact with the central molecule. Thus, a parameter f appears in the theory which increases slowly with concentration and approaches a limit corresponding to the coincidence of the shell with the wall of the nearest neighbor cage at high concentrations. In the framework of the present theory, f must be regarded as a semiempirical quantity. The hydrodynamic problem can be handled rigorously. The relative viscosity calculated is in satisfactory agreement over a range of concentrations with empirical expressions successfully applied to pertinent systems. The initial value of f derived from the observed coefficient of the quadratic term is also reasonable.
23(1952); http://dx.doi.org/10.1063/1.1702339View Description Hide Description
Using empirical relations established in the study of electron diffraction patterns of uniform thin films of aluminum, the contrast between the most intense ring and the background is calculated for other morphologies as a function of the accelerating voltage on the camera.
In all cases the contrast increases with voltage. The gain is marked if the specimen is nowhere very thick. For specimens that have very thick regions the contrast increases approximately as the square root of the voltage in the range 50 to 150 kv. The situation in reflection‐diffraction in general lies between these extremes. Increasing the voltage has a more limited effect on increasing the range of specimen thicknesses that can be examined than had been previously supposed.
Methods of Measuring the Properties of Ionized Gases at High Frequencies. III. Measurement of Discharge Admittance and Electron Density23(1952); http://dx.doi.org/10.1063/1.1702340View Description Hide Description
23(1952); http://dx.doi.org/10.1063/1.1702341View Description Hide Description
Self‐diffusion in pure polycrystallinesilver has been measured using Ag110 as a tracer. The results of the high temperature experiments, where volume diffusion is predominant, are in excellent agreement with those of other investigators. A least‐squares calculation using all available data on volume diffusion of silver gives as a value for the diffusion coefficient D=0.724e −45, 500/RT . The occurence at low temperatures of grain boundarydiffusion, in which the activity decreases exponentially with the first power of the penetration depth, as reported by Hoffman and Turnbull, is confirmed.
23(1952); http://dx.doi.org/10.1063/1.1702342View Description Hide Description
Experimental investigations show that Langmuir probe measurements can give quite erroneous results, due to work function changes of the probe, insulating layers on the probe surface, or different work functions at different places of the probe. The deposition of evaporated cathode compounds, the adsorption of oxygen released from oxide cathodes, and the sputtering and evaporation from the probe may change the work function of a pure tungsten probe in time scales of seconds. In oxide cathode tubes especially, it is necessary to measure immediately after the probe has been thoroughly cleaned. An X ‐ Y plotter, which recorded the whole probe characteristic in seconds, was found to be most suitable. No deviation from the Maxwellian velocity distribution of the plasma electrons could be found over a range of nearly four orders of magnitude in probe current, even in oxide cathode tubes under conditions where such deviations have been found by other authors.