A Repetition of the Trouton-Noble Ether Drift Experiment Citation Chase, Carl Trueblood (1926) A Repetition of the Trouton-Noble Ether Drift Experiment. Master's thesis, California Institute of Technology. doi:10.7907/7PJ3-VQ91. https://resolver.caltech.edu/CaltechETD:etd-06022004-142822 Abstract No Abstract. Item Type: Thesis (Master's thesis) Subject Keywords: Physics Degree Grantor: California Institute of Technology Division: Physics, Mathematics and Astronomy Major Option: Physics Thesis Availability: Public (worldwide access) Research Advisor(s): Unknown, Unknown Thesis Committee: Unknown, Unknown Defense Date: 26 May 1926 Record Number: CaltechETD:etd-06022004-142822 Persistent URL: https://resolver.caltech.edu/CaltechETD:etd-06022004-142822 DOI: 10.7907/7PJ3-VQ91 Default Usage Policy: No commercial reproduction, distribution, display or performance rights in this work are provided. ID Code: 2385 Collection: CaltechTHESIS Deposited By: Imported from ETD-db Deposited On: 02 Jun 2004 Last Modified: 03 Oct 2019 23:17 Thesis Files Preview PDF (Chase_ct_1926.pdf) - Final Version See Usage Policy. 354kB Repository Staff Only: item control page

A Repetition of the Trowtqn-Noble Ether Drift Experiment Master Thesis of Carl T.Chase California Inetituta of Technology May 26, 1926. A Repetition of the Trouton-Noble Ether Drift Experiment. I. Introduction. If @ parallel. plate condenser, moving with respect to a stationary ether, is charged, it can be shown that under certain conditions there will exist a couple tending to rotate the comienser. If the condenser is suspended by means of a delicate fiber, in such a manner that the plates are in a vertical plane, the couple will be given by the expression 2 Kt nev B sin2YsinZ where C is the capacity of the condenser; B the ratio of the velocity of the moving condenser to the velocity of light; V the potential to which the condenser is charged; Y the angle between the direction of motion and the plane of the plates; and 2 the angle between the direction of motion and the fiber ( the zenith distance ).The couple is a maximum when Y is 45 desrees and Z is 90 degrees. There are some who maintain that the right side of this formula should still be divided by the dielectric constant of the dielectric used in the condenser. The effect of such a factor on the sensitivity of the apparatus will be discussed later in connection with the experimental observations. Knowing the elastic constants of the suspending fiber, the rotation to be expected for a given velocity can easily be computed. II. Discussion of Previous Results. The first attempt to detect this turning of the condenser was made about I902 by Trouton and Noble who looked for the effect due to the orbital velocity of the earth combined with the proper motion of the solar system. Their experiment gave what was at that time considered a negative result; there was no effect of the size that they sought. However, in view of the recent more accurate performances of the Michelson-Morley experiment by Prof. Dayton C. Miller and the velocity of IO km/sec that seems to be cbservable, it became desirable to repeat the experiment with more sensitive apparatus and with the observations extended over a longer period of time, ( they had observed merely a few deflections each time the experiment was tried) and see just how negative their result really was. The experiment was also performed by Tomaschek in 1925 with apparatus apparently having sufficient sensitivity to detect a motion of IO km/sec and no motion was observed. However it has been found that there were serious sources of error in the apparatus that would have produced a negative result in any cuse. The arrangement consisted essen- tially of a condenser surrounded by metallic shielding and suspended on a wire of phosphor-bronze, 50 cm long and .OOI5 om in diameter.Since two connections are necessary to charge the condenser, a fine wire connected to the bottom of the condenser and dipping into a dilute solution of sulfuric acid served as the second electrical connection. A similar arrangement was set up by the author in order to determine whether the surface tension of the liquid, which was suspected of exerting a frictional force on the submerged wire, would cause any fric- tional couple comparable in size to the couple that would result from a motion in the ether. A suspended system of condenser and shielding was used having a period of oscillation of 80 seconds when hung on a wire of phoaphor~bronze, 90 cm long and .00I5 om in diameter. A fine wire attached to the bottom of the condenser dipped 2 mm below the surface of a dilute solution of salt in water. It was found that when the condenser was initially at rest, the top of the suspending wire could be turned thru two or three complete revolutions without any resulting motion of the condenser. Clearly, then, there were frictional forces in the liquid surface that would prevent any rotation of the size computed for 2a velocity in the ether of 10 km/sec (a few degrees at most). By bodily disturbing the condenset it could be made to oscillate, but only about some position of equilibrium determined by the horizontal forces exerted on the submerged wire by the surface tension, pulling the wire to one side or another depending on the initial adjustment. To remove the condenser from such a position of equilibrium it was always necessary to turn the top of the suspending wire thru several revolutions. The condenser used by Tomaschek was charged by means of a small static machine, the voltage being read on a Braun electrometer. It is difficult to see how the potential could have been kept at all constant with such an arrangenent. This would be extremely important, since rotations of the condenser caused by electrostatic forces wore always present, and could not be accurately allowed for unless the potential was kept constant. III. Deseription of Apparatus. In the present apparatus the two sources of error described above were eliminated in the following way. The lower connection to the condenser consisted of a duplicate of the upper suspension, each being of phosnhor~bronze, 90 cm long and .CO15 cm in diameter. The voltage was supplied by a generator in series with batteries, was read by means of a voltmeter, and was kept constant to a half of one percent, the limit of accuracy of the calibration of the scale of the moter, The potential was 600 volts. During the whole time that the condenser was charged it was continuously connected to the source of potential, thus preventing any change of conditions that would result if the charge leaked aWay and the potential decreased. The condenser was built up of mica discs 4 x .0C3 cm, and aluminum discs 2 x .003 cm having connecting tabs 2 mm wide. Sixty such elements were assembled in a press with the cormmecting tabs brought out on opposite sides, boiled in paraffin, and cooled under pressure. The thickness was 3 mi, the weight 10 gm, the measured capacity .04 microfarad. In the exper imental arrangement used the condenser is completely enclosed by 4 steel cylinder 23 cm long, 9 em inside diameter with walle 14 om thick. Besides serving as an excellent electrostatic shield, the larse heat capacity of the steel eliminated all disturbances from convection currents except under such extreme temperature conditions as could easily be avoided. The cylinder wes provided with caps of brass 1 cm thick. The suspensions were suppor ted and enclosed by brass tubes 1 m long and 2 om in diameter. The condenser Was closely wrapped with a layer of aluminum foil for an electrostatic shield the area of the metal plates being completely covered. This was connected to the lower terminal of the condenser, a small space being left around the upper terminal. The lower terminal was connected to the case thru the lower suspension. A smal] mirror for observing deflections was fastened to a piece of #16 aluminum wire, one end of which was securely fixed to the lower terminal of the condenser, the other end holding the lower phosthos- bronze wire. In order not to disturb the symmetry around the condenser, the mirror was Placed below the steel cylinder in the lower brass tube and could be seen thru a small hole in this tube, the hole being covered by mica to prevent the entrance of air currents. Insulation of the upper suspension was effected by a paraffin cylinder, and means were provided for adjusting the two suspensions. The period of oscillation of the condenser was 40 seconds, and when set oscil- lating it would continue to oscillate for from 20 to 30 minutes. Tous it will be seen that conditions inside the cylinder were very symmetrical. The metallic covering of the condenser , the mirror and the case were all at the sams potential, the only parte at high potential with respect to the case being the upper suspending wire and the upper terminal of the condenser (and of course the condenser plates, but these were covered completely by the foil coat). The upper brags tube was accurately concentric with the steel cylinder. In this way electrostatic disturbances were reduced to a minimun, Iv. Method of Observation. Observations were made every five minutes for twenty-four consecutive hours in the following manmer. The condenser is initially uncharged and oseil- lating. Five consecutive turning points are taken, reading tc tenths of milli- metere the position of a galvanoneter lamp image as thrown on a ground glase scale by ths mirror attached to the condenser. The soale distance was two meters. The image was always sharp and accurate readinge were easy to make. These five turning pointe were averaged to forma single "resting point". The condenser was then charged, allowed to swing for two or three minutes to overcome any possible shock or effect of @ ballistic throw caused by the sudden charging, and five more turning points observed. Care was taken to have observations equally spaced in tite so that any drift of the zero position on the scale could be accurately allowed for. Thus one resting point was obtained for each five minutes during the run, alternately with the condenser charwed and uncharged. To obtain the deflections resulting from charging the condenser, any three consecutive resting points were taken, say two for the ccndeneer uncharged and one for the condenser charged, The mean of the two uncharged positions was subtracted frem the charged position; thus any zero drift was sliminated. Tor the next deflection, two charged positions and one uncharged position were used, and so on, obtaining one deflection for every five mirutes during the rum of twenty-four hours. ‘The temperature was read once every hour. The total change in termerature was never more than one degree and often much less. Vv. Results. The deflections chserved during two runs were plotted and for comparison & curve was comeuted for the velocity that Miller obtains, and was plotted in the same diagram; the vertical distance from maximum and minimum in this curve corresponds to a deflection of 28 mm. The deflections that are at first apparent are due to electrostatic forces and depend on the initial adjustment of the apparatus, but are constant as long as these adjustments are unchanged. The downward trend of the points is due to elastic fatigue of the suspending wires. Lord Kelyin has shown that if a system is suspended by a wire and made to oscillate continuously about the axis of the wire. it becomes increasingly difficult for a given force to produce & given twist in the wire, in other words, the wire becomes stiffer. If the wire is allowed to rest for a time, it regains its original elastic froperties. In the present experiment, the condenser imakes over two thousand oscillations during a single run. The downward trend of the observed points appeared in every curve that was taken, the slope depending on the amplitude of oscillation. In one case where the amplitude was relatively large, the slope was about 30 degrees. Changes in slope always occur simultaneously with changes in temperature. There is no periodic change in the size of the deflections that is the same on any two curves. “Soretimes a $1111 maxitmmaa will occut at a riven time, sometimes a siaall1 mininum, and cften there will be no sign of either. Such maxima or minima as do occur correspond to a velocity not greater than 2 lan/sec and there is nothing periodic or systematic about these deviations. The mean of several runs is nearly a@ straight line, with no deviations corresponding to & velocity areater than 1 kn/sec. If the size of the deflections mst be divided by the dielectric constant as mentioned in the introduction, the greatest velocity that could remain undetected in any one run of twenty-four hours would de 5.5 lau/sec; and in the mean of several runs, 4 lo’sec. VI. Conclusion. Thus the experiments has been repeated with the elimination of scurces of error that have previcusly pravailed, and under more controllable conditions that have heretofore been obtained. The apparatus could detect a motion relative to the ether of 4 ku/sec, and no evidence of such a motion is found,