6.| 



NA TURK 



\May 24, 1 877 



before 1787, and employ the number of periods he has 

 hir.iself given for the interval, we find — 



I787-3-I705-5 _ Si -8 



iO'23 years. 



If, then, we commence with the epoch of 17S7 and 

 compare it with any epoch of maximum since, we shall 

 always find for the mean duration at the least 1 1 '9 years 

 according to Dr. Wolf; and if we compare it with any 

 of the epochs given by him upwards of eighty years before, 

 we shall never find a greater mean than 1075 years, and 

 this result includes an interval of 172 years before 1787, 

 with all the uncertainty of the earlier epochs. This great 

 difference of more than one year in the mean duration, as 

 derived from eighty-four years after 1787, and eighty-two 

 to 172 years before, disappears to a great extent if we 

 admit three periods iDetween 1787 and 1818. 



It has been already remarked that the duration of a 

 period is not constant, but varies within certain limits. 

 The question naturally presents itself — Does this variation 

 follow any law, or is it accidental, increasing one year and 

 diminishing the next ? The number of periods for which 

 we have the epochs of maxima and minima of the diurnal 

 oscillation of the magnetic needle accurately determined, 

 is not sufficient for any very sure reply. At the same 

 time the results I have obtained indicate a period of 

 nearly forty-two years for the repetition of the variations in 

 question ; and if this conclusion is confirmed by next maxi- 

 mum, that should occur in the year 1879. It may also be 

 pointed out that according to the law of forty-two years 

 a maximum should have occurred in 181S-42 = 1776. 

 Now this year, according to Dr. Wolf, was a year of mini- 

 mum. The variation of his sun-spot numbers for that 

 period, it appears to mc, is not sufficient to give his con- 

 clusion much weight ; while, on the other hand. Van 

 Swinden's result, which it is extremely probable was 

 a consequence of the decennial law, gives 1776 for the 

 year of maximum ; and that it was so is further sup- 

 ported by the magnetic observations of Cotte, at Mont- 

 morency. The exceptional period about 1797 shows, 

 however, that any definite conclusion from observations 

 during the last sixty years may be impossible, since causes 

 of variation exist with which we are insufficiently ac- 

 quainted as yet. 



When we compare the mean range of tlie diurnal oscil- 

 lation of the needle for the year in which it is a maximum 

 with that for the year of minimum at any station, we find 

 that the ratio of the two is very nearly constant for places 

 so widely separated as Toronto, Dublin, Trevandrum, and 

 Hobarton. I have also found that the law of the diurnal 

 movement is the same in the year for which the range is 

 least, and in that for which it is greatest. This shows 

 that it is the same cause which is acting, the variation 

 being one of intensity only. Since few or no sun-spots 

 are visible in the years of minimum range, we perceive 

 that the sun-spots happen only when the intensity of the 

 force producing the magnetic variations exceeds a given 

 value. It also appears that considerable variations in the 

 amount of magnetic disturbance may exist near the equa- 

 tor when there are few or no sun-spots ; and, on the other 

 hand, that the spotted surface of the sun may be a maxi- 

 mum, and no corresponding increase of the magnetic 

 oscillations be visible. The latter are, however, excep- 

 tional cases, since increases of sun-spots and of magnetic 

 movements occur frequently near the same time ; the 

 .'ncrease of the one, however, bears no constant propor- 

 tion to that of the other. 



It has been already stated that the ratio of the diurnal 

 oscillation of the needle in the year of maximum to that in 

 the year of minimum is very nearly constant for places very 

 widely separated from each other ; there are, however, 

 slight variations in the ratio shown at some places ; thus, 

 although it is nearly the same at Toronto, Dublin, Trevan- 

 drum, and Hobarton (I'SS), it is sligiitly greater for Munich 



and Lisbon (171). This is probably due to the action of 

 disturbances which are known to obey local laws. I have 

 also found for Trevandrum, nearly on the magnetic equa- 

 tor, that the disturbances, or the deviations of the mag- 

 netic needle from the mean position, do not show exactly 

 the same epochs of maximum and minimum in the de- 

 cennial period when different hours arc considered. Thus, 

 though the cause is cosmic, the actions appear to be in- 

 fluenced, though but slightly, by circumstances of locality. 

 When we seek for the cause of the decennial period, 

 we are met at first by the three phenomena which obey 

 this law : the magnetic variations, the sun-spots, and the 

 aurora borealis. The connection between the first and 

 third is so marked, that if a magnetic disturbance com- 

 mences during the day in a high latitude, it is quite cer- 

 tain that the aurora will be seen as soon as the disappear- 

 ance of sunlight permits. This is a fact I have verified 

 during several years' observations in the south of Scotland. 

 Both these phenomena are results of electrical motions. 

 It did not seem improbable then that the solar spots 

 might be connected with disturbances of electrical equi- 

 librium, and that these might be due to the different 

 electrical states of the sun and of the planets. 



We do not know, however, of any planet with a period 

 of ten and a half years, nor of any combination of planetary 

 positions which would produce such a period. My own 

 researches have failed in connecting the variations of the 

 sun's spotted surface with the time of revolution of any 

 planet by a law which holds for different decennial 

 periods. This fact, however, does not disprove a planetary 

 action. We are unacquainted with the nature of the 

 medium through which the electrical actions producing 

 the magnetic variations are conveyed. Physicists seek 

 to reduce the phenomena of nature to the fewest possible 

 factors : many then have been induced to believe that 

 electrical and magnetical actions are conveyed by the 

 same ethereal medium which we believe transmits heat 

 and light. The facts do not appear to be easily ex- 

 plained by such a hypothesis ; thus I have found that 

 certain electrical actions of the sun producing marked 

 diminutions of the earth's magnetic force happen exactly 

 at successive intervals of twenty-six days ; when one point 

 or meridian of the sun returns to the same position rela- 

 tively to the earth ; this action, similar to that of a beam 

 of light reflected from a revolving mirror, which illuminates 

 a particular point only at the same part of its revolution, 

 has no resemblance to that of light and heat, which are 

 propagated equally in all directions. 



If, then, we can suppose that the electrical medium is 

 disposed unsymmetrically around the sun, that the dis- 

 position and extension varies, it is obvious that the sup- 

 posed planetary actions would also vary, and might be 

 quite different for different parts of their orbits, m dif- 

 ferent decennial periods. This suggestion may explain 

 why I have not been able to find a law remaining the 

 same in the different periods ; and it is not opposed to 

 the conclusions of Messrs. De la Rue, Stewart, and 

 Loewy, who have found very remarkable relations be- 

 tween certain positions of the planets and the amount of 

 the sun's spotted surface during a single decennial 

 period 



Any hypothesis which seeks to explain the mode of 

 production of the sun-spots (by cyclones or otherwise) 

 must also explain why the causes become insulticient for 

 their production every ten and a half years. M. Faye, 

 the distinguished French astronomer, considers that the 

 prime cause of sun-spots is to be found in the excess of 

 heat radiated ; so that the spots are the symptoms of a 

 dying sun ; that we have in fact here a phenomenon like 

 the flickering of an expiring lamp which may have a 

 periodical character. Such a hypothesis will scarcely 

 satisfy the demands of science, but we must evidently 

 wait for more facts before any satisfactory theory can be 

 proposed. John Allan Broun 



