The experiment described by ]VTr.
Nichols in the foregoing article should be of particular interest toevery-one who has followed the development of precision frequency measurements during the past few years. The crystal-controlled standard-frequency assembly was of course designed and built primarily for measuring frequencies anywhere in the communication-frequency spectrum and its direct application to a time measuring problem serves to emphasize the identity of time and frequency.
Reduced to its lowest terms the time measuring problem in the Michelson experiment is nothing more than the determination of the rotational frequency of the rotating mirror. The use of the standard-frequency assembly to measure without intermediate steps this rotational frequency is a possibility that will suggest itself to most of us who are accustomed to think in terms of frequency rather than in terms of time. It becomes a problem of measuring a frequency of some 500 cps. with the greatest possible accuracy, say, one part in three million. If the tuning fork used in the Michelson experiment could be replaced by a frequency derived from a standard-frequency assembly, the question of determining this rotational frequency in terms of a standard time interval would be considerably simplified.
With this simplification in mind it may be of interest to consider a suggestion made several years ago by Major William Bowie of the United States
This is how the jVlichelson experiment might, so to speak, be inverted to measure distance in terms of the velocity of light and time. The rotational frequency of the mirror would be measured by a standard-frequency assembly, thus yielding the length of time required for the mirror to move an adjacent face into position
Coast and Geodetic Survey: that distance be measured in terms of the known velocity of light and time. A measurement of this kind would have important practical applications in geodesy where the difficulties of laying down precise base lines in mountainous countries and in archipelagoes are serious when done by present methods.
The experimental basis for such a measurement of distance was laid by an earlier experiment of Dr. Michelson when he measured the velocity of light between two stations — one located on Mount Wilson and the other on Mount San Antonio 22 miles away. The distance between these two stations was measured by the United States Coast and Geodetic Survey with a probable error of one part in 6,800,000 and the time was measured by methods somewhat similar to those just described by Mr. Nichols.
In measuring distance the experimental procedure would be reversed, and, instead of measuring the time it takes light to traverse a known distance, we would measure the time taken by light to traverse the unknown distance and then work out the distance from the known velocity of light. The compactness, portability, and high precision of a standard-frequency assembly would readily adapt it to this use since the elimination of as much bulky apparatus as possible would be desirable from the experimenter's point of view.
Some time we hope to see the experiment tried. It certainly has many interesting possibilities.
Was this article helpful?