On the periodicity of sun-spots

In the year 1906 I presented to the Royal Society an investigation in which it was shown that the frequency of sun-spots was subject to recurrent variations, not only in the well-known 11 years’ cycle, but also in other periods, which were determined. As we now possess additional material extending over 10 years, it is interesting to examine how far the minor maxima of the subsidiary periods can be traced in the more recent records. The accompanying figure gives diagrammatically the sun-spot areas measured at Greenwich between 1898 and 1909. The numbers plotted represent the sum of the mean areas during four successive rotations, beginning with the four rotations 593/596 of Carrington’s series. There is a period of 4.79 years, which in the previous communication was shown to be persistent during the whole time covered by sun-spot records, more persistent, in fact, than that of 11 years. I have marked on the diagram the predicted times of maxima of the period with an arrow pointing upwards. The first maximum, towards the end of the year 1898, which was timed to arrive during the rotation 604, actually took place three rotations, or about months, earlier. The second maximum (September, 1903) was predicted to take place during rotation 668. It actually took place during the rotation 671, though an almost equally strong maximum was observed during rotation 667. We may therefore say that there is here an almost absolute coincidence in the predicted and observed times. The maximum of July, 1908, was delayed by about two months, but the activity had already risen considerably at the predicted time. In all three cases the coincidences of the predicted and actual times are very satisfactory, if it be remembered how variable is the observed maximum of the 11 years’ period. This periodicity of 479 years seems characterised by one or two sharp outbreaks near the time of the maximum, and throughout the time that accurate records are available it nearly always shows itself in each cycle as a separate peak in the curve representing sun-spot frequencies. The outbreaks of sun-spots connected with this period can be traced also in the magnetic records. There were several disturbances during September and October, 1898, notably one on September 9. In 1903 there was a magnetic storm on October 12, and more violent ones at the end of the month. Finally, in 1908, we had strong distxirbances on September 11 and 29.