Volume 4, Issue 9, 01 September 1933
Index of content:
4(1933); http://dx.doi.org/10.1063/1.1745196View Description Hide Description
In experiments with a Gill‐Morrell oscillator generating undamped waves less than a meter in length, a study was made of the potential difference between the Lecher wires coupled to the oscillator by enclosing these wires in an evacuated glass tube several wave‐lengths long. The oscillator itself employed two type −52 tubes connected in push‐pull and it set up radiofrequency currents of 2.5 amperes in the Lecher wires and potential differences in excess of several hundred volts between the wires at positions separated by half wave‐length intervals. At these voltage loops for certain critical pressures, between the wires a discharge took place which was photographed as well as employed in a direct study of the standing waves in the wires. The gases used in the tube were air, helium, oxygen, mercury vapor and mixtures of these gases with air. The use of this tube made possible a visual study of the changes in potential distribution as various circuit adjustments were made. A series of double maxima was observed as the length of the line was changed by a half wave‐length.
4(1933); http://dx.doi.org/10.1063/1.1745197View Description Hide Description
Ions formed by collision in electronic amplifier tubes move toward the cathode and control‐grid and produce momentary increases in the space‐charge limited current (current pulses). These current pulses are random in time and produce fluctuation noise analogous to the shot‐effect. An approximate theory is presented, applicable to cathodes of large diameter (ca. 0.1 cm). Both the noise and the increase in the average value of the electron current due to the presence of ionization are calculated in terms of the ``size of the elementary event'' or the integral of the current pulse. It is not feasible to calculate the latter directly but it may be evaluated indirectly by measurements of the increase of average current. The noise calculated from the values so obtained and from the ionization probability data of Smith, Bleakney and Tate is in close agreement with that observed experimentally. The production of noise in tubes with oxide cathodes containing mercury vapor, argon and the gases naturally evolved from the electrodes and tube walls is investigated experimentally as affected by pressure and electron current density. In the case of a tetrode containing mercury vapor or evolved gases the variation with pressure is linear up to 10−3 mm; in the case of argon the noise varies as the 1.1 power of the pressure. In the mercury vapor tetrode a variation as the 3/2 power of the plate current was found, as compared to a 5/3 law, expected theoretically. These measurements were made at 620 kilocycles. The noise per frequency interval was measured as a function of frequency over the range 500–1500 k.c. At pressures of the order of 10−4 mm in mercury vapor a decrease at the higher frequencies was observed; with argon the noise was uniform. With mercury vapor at higher pressures (ca. 4×10−3) the noise spectrum became peaked, the frequencies of the peaks depending upon the electron current flowing (electrode potentials). The hypothesis that these peaks are due to oscillations of positive ions in the potential trough surrounding the cathode is in accord with the principal experimental facts. With further increase in pressure continuous oscillations are produced, sustained apparently by a regenerative action of the space charge upon the oscillating ions. The frequencies of these oscillations are affected only by the electrode potentials and not at all by the external electrical circuit.
4(1933); http://dx.doi.org/10.1063/1.1745198View Description Hide Description
The direct problem of deducing from electrical potentials observed at the surface of a horizontally uniform earth the unknown variation of the conductivity with depth reduces to a boundary value problem of unusual type. Its solution for the isotropic case is developed in Part I, following. In Part II the more usual inverse problem is solved for some special classes of conductivity functions and graphical examples are given as an aid in guiding the interpretation work. In Part III, the anisotropic case is discussed.
4(1933); http://dx.doi.org/10.1063/1.1745199View Description Hide Description
The formula for the mutual inductance of two concentric solenoids which is used probably more than any other is the one originally given by Maxwell. Since the solenoids were taken to have negligible thickness in the above formula and since the effect of the thickness is often appreciable in practical cases, additional terms to give the effect of the thickness have been calculated and are presented in this article.
4(1933); http://dx.doi.org/10.1063/1.1745200View Description Hide Description
Three different types of low voltage circuits suitable for supplying impulse voltages in oscillographic work are discussed. These circuits use Thyratrons as the connecting and timing switches in place of the spark gaps ordinarily employed in transient investigations. Charged condensers are used as the source of voltage. Two circuits are of the single tube type and the third is a multiple tube system in which a number of condensers are charged in multiple and discharged in series by means of Thyratrons. Oscillograms of the voltage form obtainable in each case are given and the time lag of starting determined.
4(1933); http://dx.doi.org/10.1063/1.1745201View Description Hide Description
Thermal conductivity of Acheson graphite is reported for the temperature range −150°C to +700°C. Cylindrical blocks of the material were placed above and below a flat electric heater and lateral losses balanced out by auxiliary coils wound on the surface of the sample. The results show a continual decrease in thermal conductivity with temperature over the entire range. The law, k/aC=A/T+B, proposed by Bidwell for certain other materials is found to hold accurately for graphite.
4(1933); http://dx.doi.org/10.1063/1.1745202View Description Hide Description
The psychrometric constant ψ, defined as the ratio of A (corrected for radiation) to the A predicted by the convection theory, is shown to be a function of velocity. At infinite velocity, the new equation reduces to that of August (ψ=1); at zero velocity, to the result obtained by Maxwell, ψ=K/D. An improved method of obtaining the true wet‐bulb temperature is given, together with values of ψ, accurate to about 1 percent, for toluene, chlorbenzene and xylene. Straight lines are obtained when ψ for the latter two liquids against ψ for toluene at the same velocity. The theoretical psychrometric line for water is found; it is shown from theory that the line uncorrected for radiation possesses a stationary minimum value at velocities between 1000 and 10,000 ft./min., the radiation error just balancing the deviation from the August theory. In this range A is constant at 630×10−6, while at lower velocities it varies rapidly because of the change in the radiation error. For accurately reproducible results, the air velocity must be maintained at 1000 ft./min. or more.
4(1933); http://dx.doi.org/10.1063/1.1745203View Description Hide Description
It has been shown previously by the author that, from the physical point of view, the process of propagation of the nerve impulse is essentially different from the propagation of other kinds of disturbances usually studied in physics. Instead of being described by a differential equation, this type of propagation leads to a simple integral equation. In the domain of the inorganic similar types of propagation are met in the spread of activation on the surface of passive metals. In the present paper the problem of such types of propagation is treated mathematically for two different cases. In the first case it is assumed, that the nerve is electrically uniform all along its length. In this case the final formula for the velocity of propagation reduces to a rather simple expression which applied to the ischiadicus of the frog, leads to a value of 15 meters per second for the velocity of propagation. In the second case the nerve sheath is assumed to be completely insulating, except at the Ranvier nodes, where its continuity is broken, so that the electrical properties of the nerve vary periodically along its length. For the second case a more complicated formula is obtained, which reduces to the first one, when the distance between the nodes tends to zero. Effects of possible distributed capacity are briefly discussed.