no 



KNOWLEDGE 



[May 2, 1892, 



«artb on its axis, the yearly revolution round the sun, are 

 both distinctly recognised by the magnets, and, if necessary, 

 we could determine the exact period of either by magnetic 

 observations alone, if carried on for a sufficient length of 

 time. 



But the third relation is of a different order altogether. 

 It is that to which I referred in the first part of this paper 

 when I said that " when we take average results for suc- 

 cessive years we find that this motion," /.<■., the diurnal 

 range, " is greater in amplitude and force in strict propor- 

 tion to the number and size of the spots upon the sun." 



It is now more than forty years since the occurrence of 

 a regular periodicity in the numbers and dimensions of 

 Sunspots was established by the observations of Schwabe, 

 a periodicity which, though now so familiar to us as a fact, 

 still remains as much as ever a mystery as to its cause. 

 .Tust before Schwabe published his great discovery Lamont 

 had detected a similar periodicity in the oscillations of the 

 magnetic needle, and directly Schwabe's Sunspot results 

 were published General Sabine and several other observers 

 of magnetic phenomena at once recognised that the two 

 periods were not only similar but identical. 



This identity did not consist in a mere equality of the 

 average period, or iu an occasional correspondence of 

 minima and maxima. It was far more precise. The attempt 

 was early made to exjjlain the Sunspot cycle as due to the 

 influence of Jupiter, the period of revolution of that planet 

 nearly corresponding to the mean solar cycle. But the 

 period of -Tupiter is a constant, whilst the solar cycle is 

 irregular in length, often varying considerably from its 

 mean value. And the collation of earlier Sunspot records 

 soon showed that, though a fair correspondence might be 

 made out between the planet's position and the numbers 

 of spots seen during two or three periods, yet if we went a 

 little further back the two got hopelessly out of step. There 

 was, therefore, no connection there. 



But the correspondence with the magnetic cycle was 

 precise. Was the interval between one minimum and the 

 next shorter than usual with the Sunspots ? Then it was 

 so also for the magnetic diurnal range. Was the spot 

 maximum delayed? So was the magnetic. The two 

 cycles have never failed to correspond in their general 

 features since both were first under regular observation. 



So far is this the case that Dr. R. Wolf, one of the first 

 of Sunspot observers, has worked out a very simple formula 

 by which he is able to convert the yearly means of what he 

 terms his " relative numbers " for Sunspots, so as to repre- 

 sent the yearly means for the diurnal range of magnetic 

 declination as observed at Milan, and a similar formula 

 would serve for other localities. 



The two cycles correspond also in more than their 

 general features or annual means. 



The life of a Sunspot being almost always short, and 

 spot groups varying very much as to size and being very 

 irregularly distributed over the sun, it follows necessarily 

 that even at a time of maximum there are very great 

 fluctuations in the proportion of the sun's visible hemi- 

 sphere which is given up to spots. So that the rise to 

 maximum, or the fall to minimum, does not proceed 

 smoothly, but in a succession of waves as it were. Of these 

 minor oscillations, any one which has been fairly well 

 marked and at all sustained is sure to have its counterpart 

 in the activity of the magnets. 



It would almost seem a necessary corollary of this last 

 circumstance, that particular and individual Sunspots 

 should he attended by corresponding magnetic disturbances. 

 It certainly was the case with the great February spot, as 

 we have but to glance at the accompanying plate to see. 

 At about half-past five in the morning of February 13th, 



the trace both of the declination and of the horizontal 

 force magnet sufl'ered a sharp and sudden deflection, token 

 of an instantaneous disturbance. This sharp and sudden 

 twitch, which occurred, so far as we know, to the needles 

 in all the magnetic observatories of the world and at 

 practically the same actual (not local) time everywhere, 

 is pre-eminently typical of the intense magnetic storms. 

 An inspection of the trace shows that its later wanderings 

 were far more considerable than this first one, but it is its 

 instantaneousness which gives it its special character. 

 In the afternoon of February 13th, the magnets were 

 vibrating to such an extent that the two traces have 

 become intermixed. Later, about midnight on February 

 13th, the magnets twitched so violently that for about an 

 hour the trace went quite ofl' the prepared paper. The motion 

 in declination that was registered amounted to more than 

 1°. The horizontal force trace went off the sheet for an 

 hour and a half, and in vertical force (not included in the 

 plate) the disturbance was so great that the trace was lost 

 in one direction, that of increasing force, for four and a 

 half hours, and in the other for an hour and a half. A 

 period of short, quick oscillations was then set up and 

 continued for some hours, the disturbance dying away m 

 the course of th€ evening of February 14th — a little later, 

 that is to say, than the trace is shown on the plate. 



So remarkable a magnetic storm, occurring just about 

 the time when so great a spot was about attaining its 

 fullest development, might well be taken as proving a 

 special connection between the two, especially when the 

 facts already alluded to with regard to the general corre- 

 spondence between Sunspots and terrestrial magnetism are 

 borne in mind. And the conclusion would be further 

 supported by the occurrence of a fine aurora on that 

 same night, February 13th, for the variation in number of 

 auroriB has also been shown to take place precisely in the 

 same cycle as the variation in number of Sunspots. 



But the proverb tells us " One swallow does not make a 

 summer," and neither can we take one coincidence as 

 proving a real connection. Since the spot group of 

 February, 1892, was the largest ever observed at 

 Greenwich, let us look through the records for the group 

 second in size, and see if that has any information to 

 give us. 



This second spot v.-e find in that of November, 1882. 

 This group was seen during three successive rotations of 

 the sun. It formed in the visible hemisphere on October 

 20th, increasing in size on the succeeding days with 

 marvellous rapidity. It passed out of sight at the west 

 limb on October 28th, and was next seen at the east limb 

 on November 12th. Its area on November 13th, when 

 the whole of it was visible, amounted to 1090 millions of 

 square miles, and this area increased day after day until 

 November 18th, when it reached the central meridian and 

 attained its greatest dimensions, having an area of 2860 

 millions of square miles. After this it began to break up 

 and to diminish again. On November 2::ird its area was 

 2030 millions of square miles. On November 2.5th it had 

 reached the west limb. It was seen at the east limb again 

 on December 10th, its area reduced now to 515 millions, 

 and by December 21st it had completely faded out. 



What was the record of the magnets at the time when 

 this spot was at its greatest development '.' This is what 

 the Astronomer-Royal chronicles : "A remarkable magnetic 

 storm, preceded by several days of considerable magnetic 

 disturbance, was observed here on November 17tli. It 

 commenced suddenly November IGth, 22h. 15m. G. M. T. 

 ■with a great decrease in all the magnetic elements, the 

 declination being diminished by more than 1^, the 

 horizontal force by more than 1 -100th part, and the 



