OX SUNSPOTS AND TERRESTRIAL PHENOMENA; 451 



from a memoir by Fritz, will repay a careful study. While especially in 

 curve 4, which takes in the largest number of observations, we can trace 

 the rise at each spot maximum, it will be seen that the amplitude of the 

 oscillation is very irregular, but suggests the existence of a secular period. 

 The length of that secular period can be approximately estimated by 

 taking account of some older observations. In this way Fritz has 

 fixed its period at fifty-five years, but this is a point on which it is safe 

 to speak with caution. 



Curve 1 of the plate represents the sunspot frequency. In curve 2 

 the European observations of aurora? between latitudes 46° and 55° were 

 taken into account. Curve 3 represents the frequency of aurora? observed 

 in America between latitudes 0° and 60°. Finally, curve 4 takes in all 

 European observations extending to the arctic circle. 



General Discussion. 



A very remarkable connection between sunspots and terrestrial 

 phenomena is thus established, and the question naturally arises whether 

 we can form any idea as to the nature of the connection. Though this 

 may seem hopeless at present, we can at least form an opinion as to how 

 the connection cannot be explained. In the first place, we must not think 

 of a direct magnetic effect between sun and earth. The sun is too hot 

 to be bodily magnetised ; and even were it made of hard steel magnetised 

 to saturation, it would not sensibly affect the magnetic needles of tho 

 earth. It has been suggested that the sun may be electrified to a high 

 potential, and by its induction affect the magnetic elements of the earth. 

 It has also been maintained that the space between the sun and earth is 

 a good conductor of electricity, and must thus bring their electric states 

 into connection. I reserve for another opportunity the discussion of 

 the possibility or probability of either of these hypotheses ; at present it 

 will be sufficient to say that I am unable to conceive how, even granting 

 "their truth, they can possibly explain the increase in the diurnal variation 

 at times of many spots. 



We do not at present know what the daily variation of the magnetic 

 needle is due to, and therefore it seems premature to consider what 

 causes may modify the effect. The solar influence bears on it the stamp 

 of a temperature effect, while the lunar influence seems tidal. It is 

 difficult to say, without very careful consideration, what the thermo- 

 electric effects in the earth's crust might be. The electro-motive force 

 no doubt is very small, but then the resistance is small also, and the 

 current might be quite appreciable. Tides in the earth's interior, like those 

 assumed by Professor W. G. Adams, 1 do not apparently exist to any appre- 

 ciable degree, but there must be a tidal stress, and though very small 

 the consequent strain might produce a difference of resistance in different 

 directions, and thus alter the channels of the electric currents. The 

 superposition of such a thermal and tidal effect seems to me to be the most 

 hopeful hypothesis at present, but I have not followed it out sufficiently 

 to be able to do more than to throw it out as a suggestion. These and other 

 considerations, however, lead us on to inquire whether these magnetic 

 and electric effects are not perhaps only secondary, and that the sun acts 



1 Royal Institution, June 3, 1881. 



G Q 2 



