Volume 13, Issue 7, 01 July 1942
Index of content:
13(1942); http://dx.doi.org/10.1063/1.1714890View Description Hide Description
A simple method is described for making replicas of surfaces for study in the electron microscope. The surface to be studied is cleaned and then coated with a very thin slightly wedge‐shaped film of polyvinyl formal having a range in thickness from about 500 to 750A in 2.5 cm. This film is formed by dipping the prepared surface into a 0.5‐percent solution of the resin dissolved in dioxane or ethylene dichloride and permitting the solvent to evaporate while the surface is held in a vertical position. The replica film is removed by placing the coated surface under water and peeling the film from it. A description of this technique is given in detail. A graph is presented which shows the thickness of resin films obtainable from various concentrations of resin in dioxane, and a table gives the interference effects which may be used for determining approximate thicknesses of thin resin films. Several easy procedures are described for checking the thickness of such films. A technique is described for preparing strong films having a thickness of 100A and an area of 3 cm2 when mounted in air. Such films are very useful when working with the electron microscope and several methods are described for mounting specimens on them, including a conditioning treatment—useful in mounting water suspensions—which makes one surface very hydrophilic. Five micrographs show the results obtainable with the techniques described.
13(1942); http://dx.doi.org/10.1063/1.1714891View Description Hide Description
The general case of vibration of a shaft with several rotating masses is discussed. The analysis is reduced to a simple matrix multiplication of second‐order matrices and the natural frequencies of symmetrical systems are discussed. The connection of the matrix method and the well‐known Holzer method of computing the natural frequencies of torsional oscillations is considered. The analysis of geared systems and continuous shafts is discussed.
13(1942); http://dx.doi.org/10.1063/1.1714892View Description Hide Description
Meek's criterion for the formation of a sparkdischarge in a gas is criticized. An error in the derivation of his formula for the radial field about an electronavalanche channel is noted, and the value adopted for this field is shown to be too small to draw the postulated number of electrons into the channel from the space outside. The importance attached to the radial field is found to be misplaced because of the above and because a criterion based solely on the total number of ions produced in an avalanche gives results for sparking potentials for different spark lengths and for a gas at different pressures in agreement with those obtained by Meek's method. Both methods give results which, using the values assigned to the constants, fall more and more below the experimental values of the sparking potentials as the length of the spark gap, δ, is increased until at 20 cm the deviation is over 7 percent. To bring the computed value for δ=20 cm into agreement with the experimental value the constants in the equations would have to be increased by a factor 2×1016 over the values which give agreement between formulas and experiment for δ=1 cm. The dominant exponential term in each formula is based on the incorrect assumption that the field is not affected by the space charges in the avalanche. That a large amount of ionization, postulated by Loeb and Meek, is produced by a high potential wave which travels through the channel as soon as the positive streamer reaches the cathode is questioned because the fields found in such waves are too small to produce ionization by electron collision in air at atmospheric pressure.
13(1942); http://dx.doi.org/10.1063/1.1714893View Description Hide Description
The Hurter and Driffield curves and the absolute sensitivities of Eastman medium lantern slide plates are determined for electrons accelerated through differences of potential from 40 to 212 kilovolts. Within experimental error the H and D curves are identical in shape for the different voltages. The sensitivity reaches a peak near 100 kv and then drops off rapidly. This phenomenon may be utilized to increase the effective sensitivity of plates at high voltages by applying a thin, high density intensifying screen to the surface of the emulsion.
13(1942); http://dx.doi.org/10.1063/1.1714894View Description Hide Description
Saint‐Venant's torsion problem is solved for the linear prism with right cross section bounded by the curve, an inverse of an ellipse with respect to its center. By a method developed recently by R. M. Morris, explicit formulas are obtained for the torsion, conjugate torsion, and complex torsion functions; for the twisting moment; for the stress function and the components of shearing stress. This work is of interest as complementing recent work by A. C. Stevenson and by D. L. Holl and D. H. Rock; further, as an addition to the very few exact solutions that have been effected for cross sections bounded by single, closed, continuous, partially concave curves.
13(1942); http://dx.doi.org/10.1063/1.1714895View Description Hide Description
It is proved that a human being M falling from a ledge A through the air to the ground below will attain the greatest range if M falls forward while standing upright. If M was found on the ground at a position B beyond this range of fall, M furnished energy himself to attain B. An analysis of trajectories from A to the ground due to an effort on the part of M is made. If the kinetic energy which M must produce to reach B is much greater than that which M can furnish after loss of balance, M made the trip from A to B by a premeditated jump. Such analyses are important for determining suicides.