Periodical variation of the velocity of light.
236
In No. 5520, the writer has pointed out that observationirl evidence conclusively shows that the velocity of light was decreasing during the last quarter of the last century, and that it was still decreasing in the present century, the rate being approximately 4 km/sec per year.
This was confirmed by the work of KaroZus and MitteZstaedt in 1928, who obtained a value of 299778f20 for this pseudo-constant, using a Kerr cell to interrupt the luminous beam, on a base of 41.386 metre (Yhysikalische Zeitschrift 1928, p. 698). The last va.lue obtained before this was MzcheZson's, who got 299796f4 over a base of 35385.53 metre, by the revolving mirror method; the decrease was therefore 18 km/sec a little smaller than the probable error of the measurements.
))Nature(( (February 3' " 1934, p. 169) reports that Pease and Pearson, at the Mount Wilson's observatory, have obtained 299774 km/sec in 1933, with a probable error of 1 or 2 km/scr. This is again lower than that of KaroZm and MitteZstaea't, fully bearing out the conclusion arrived to bv the writer, that the velocity of light is decreasing.
This last value, however, occupies such a position on the graph as to justify the writer taking up again the suggestion he made in 1927 (L'Astronomic, November, p. 504) that the variation may be a fluctuation of the velocity of light, in which case both the decrease observed in the last century and that shown to take place by the determinations made since 1920 may have been separated by an intermediate period during which thc velocity increased, but which, unfortunately, is only represented by one determination, that of Perrotinin 1902 made by the toothed wheelmethod.