Interferometric measurements in the spectrum of krypton

Abstract

The standard of length at present accepted by spectroscopists is the wavelength of the red line of cadmium, 6438·4696 A., in dry air at N. T. P. This defines the angstrom, and although it probably differs from 10 -10 metre by little more than one part in ten millions, it seems desirable to retain the present definition, since there is now some doubt as to the permanency of material standards of length. Moreover the present work shows that it is quite possible to compare wave-lengths with an accuracy of at least one part in fifty millions, or probably about ten times the accuracy which can be attained in the comparison of the lengths of a material standard and a light wave. Unfortunately the red line of cadmium is not very convenient as a spectroscopic standard of length. It is situated in the deep red, is therefore somewhat difficult to photograph, and it is one of the fainter lines in the spectrum of cadmium. The Michelson lamp, which is the only strictly correct source of the radiation, requires to be heated in a furnace at a temperature of 320° C. This is highly undesirable since the heat from the furnace greatly increases the difficulty of making accurate interferometric wave-length comparisons. Finally the line is in a spectral region far removed from that where the majority of spectral lines occur, which is disadvantageous, since it is very much easier to make accurate comparisons of wave-lengths not far removed from each other.