The relative frequency stability of two beamtype maser oscillators is used to test the dependence of the velocity of light on velocity of the frame of reference with considerably more precision than has been obtained from experiments of the Michelson-Morley 1 type. Expressed in terms of an ether, the maximum ether drift is shown to be less than 1/1000 of the earth's orbital velocity.The experiment, which was performed at the Watson Laboratory, involves comparison of the frequencies of two masers 2 having their beams of NH 3 molecules traveling in opposite directions. Miller 3 has analyzed this case and given the change in frequency of a beam-type maser due to ether drift, assuming the molecules in the beam to have a velocity u with respect to the cavity through which they pass, and the cavity to have a velocity v with respect to the ether. The shift may be simply discussed by assuming that, if v is zero, radiation is emitted perpendicularly to the molecular velocity so that there is no Doppler shift. If the cavity and beam are then transported at velocity v through the ether in a direction parallel to u, radiation must be emitted by the molecules slightly forward at an angle 6 =7r/2 -v/c with respect to u. The fractional change in frequency due to the Doppler effect is then e = u/c cos0 or uv/c 2 due to motion through the ether, assuming that the proper molecular frequencies are unchanged by such motion 0 For a thermal molecular velocity of 0.6 km/sec and for the earth's' orbital velocity (30 km/sec), € -2 x 10 ~1 0 . The difference in frequency due to the above effect between two masers with oppositely directed beams would be 2ev, or about 10 cps for v equal to 23 870 Mc/sec, the NH 3 inversion frequency.Although uv/c 2 is of second order in the velocities, it is of first order in the velocity of the cavity, or of the laboratory, with respect to the ether Q The present experiment measures the entire effect with a rather small fractional error, which affords a particularly small upper limit to v since this quantity enters in first order, rather than in second order as in the Michelson-Morley experiment. A somewhat similar term would occur in the latter experiment if the interferometer used were transported by a plane of speed u, and interference fringes were compared for two opposite directions of flight.Two maser oscillators with oppositely directed beams were mounted with necessary auxiliary equipment on a rack which could be rotated about a vertical axis. The beat frequency between the two oscillators was adjusted to about 20 cps and recorded continuously. After approximately one minute of recording with the maser axes oriented in an east-west direction, the apparatus was rotated 180° and the beat frequency recorded in the new position.The change in beat frequency, on the basis of an ether drift, should be 4ev, or about 20 cps. Sixteen such comparisons were made during a period of about 20 minutes. These were repeated about once per hour during a time somewhat longer than 12 hours, so that the earth's rotation...

No abstract

No abstract

No abstract