The valve maintained tuning fork as a primary standard of frequency

In 1922 an investigation was carried out at the National Physical Laboratory to determine the constancy of frequency that could be expected from a valve maintained tuning fork. It was found that the fork was capable of operating with a degree of steadiness of frequency which was greater than was then necessary for most purposes. The investigation resulted in the design of a 1000 cycles per second fork which served as the Laboratory frequency standard until 1931. For precision work it was necessary to measure the frequency of the fork during the observations by comparison with a standard Shortt clock ; but if the accuracy required was less than 2 parts in 105 it was sufficient to apply a correction for temperature to the nominal value of the fork frequency. With the rapid advance in radio frequency technique and the ever-increasing number of wireless transmitting stations the problem of frequency standardization became increasingly important; and it was decided to instal a standard, which should be in continuous operation at a frequency within I part in 106 of its normal value. As the most suitable frequency for use in conjunction with the existing equipment for the measurement of radio frequencies was 1000 cycles per second, and as the tuning fork had hitherto given a satisfactory performance, it was decided to continue the investigation on the fork to determine whether it could form a frequency standard of the desired degree of accuracy.