The relationship between the “solar constant” and terrestrial magnetism

1. The issue of an authoritative list of provisional values of the “solar constant” as determined by Dr. C. G. Abbot and colleagues, for the period August, 1920, to November, 1924, is of interest to all who concern themselves with solar radiation and its relation to terrestrial phenomena. What interests myself is the possible connection between fluctuations of the “solar constant” and the phenomena of Terrestrial Magnetism. As is well known, the solar 11-year period is prominent in the range of the regular diurnal variation of the earth’s magnetic field, and the sun’s rotation period—approximately 27 days in the sunspot zone—is recognisable in magnetic disturbance. Owing to the large annual variation in the range of the magnetic elements, conclusive evidence of a parallelism between the amplitude of that range and the size of the “solar constant” can hardly be hoped for from so short a period as four years. But the 27-day interval is usually easily recognisable in the magnetic data from a single year, and if there is any at all intimate relation between the value of the “solar constant” and the earth’s magnetic condition, either on the same day or on a definite subsequent day—i. e .., a day 1 ... , 5 or 10 days later—then we should expect a 27 day interval to show itself in Abbot’s figures. The data now published seem ample to check this point, if treated in the way which has proved successful in the case of magnetic phenomena and sunspot numbers or areas. We start by taking for each month a certain number of days selected as those in which the element under investigation has its largest values, and an equal number selected as the days in which the aforesaid element has its smallest values. We then compare the mean value of the element for two associated groups of days. The days of the one associated group follow the days of the first primary group of selected days by, say, 26, 27 and 28 days. The days of the second associated group similarly follow the days of the second primary group of selected days. If the representative day in either of the primary groups be described as dayn , then the representative days in the associated groups will be respectivelyn + 26,n + 27 andn + 28. The so-called 27-day interval is not a sharply defined interval, such as 27.00 days. For one thing, magnetic storms usually extend to more than one (Greenwich) day, and solar phenomena have also their indefiniteness. Thus it is desirable to extend the investigation to at least the two adjacent days.