48 ANNUAL RECORD OF SCIENCE AND INDlTSTRY. 



the polar light is a periodical phenomenon, whose most im- 

 portant i^eriod embraces fifty-five years, approximately. Four 

 of these periods, apparently, group themselves together and 

 form a greater period of over two hundred years, while the 

 fifty-five-year period itself is subdivided into five secondary 

 periods of about eleven years each. The fifty-five-year pe- 

 riod stands in an intimate relation to the period of the sun's 

 spots, the maxima and the minima of each corresponding 

 exactly; but the principal maxima of the auroras are more 

 decided than those of the sun's spots. The aurora stands in 

 an intimate relation to the terrestrial magnetism and its va- 

 riations : from observations gathered from the most various 

 regions of the earth, we know that magnetic disturbances 

 and auroras frequently happen at the same time, or follow 

 each other in close connection ; but the disturbances of the 

 terrestrial magnetism that accompany the polar lights are 

 not ^perceptible at all the magnetic stations, and probably 

 auroras sometimes form without being connected with any 

 disturbance of the earth's magnetism. A comparison of the 

 curves representing the changes in the various elements of the 

 earth's magnetism with the auroral curve shows that the max- 

 imum of the aurora that occurs daily about ten o'clock in the 

 evening agrees completely with that minimum of the mag- 

 netic declination which, like the maximum of the auroras, is 

 delayed the farther north we go. This aurora maximum only 

 coincides with the secondary maximum of magnetic perturba- 

 tions that is shown at many stations on the globe, because the 

 aurora maximum corresponds in high latitudes with that max- 

 imum of the magnetic disturbance that occurs after midnight. 

 The annual periods of terrestrial magnetism and of auroras 

 agree still better, since the separate elements of both agree in 

 their principal features with the equinoxes. Since the secular 

 period of the auroras agrees with the period of the solar spots, 

 and these latter again run jDarallel to the variation of mag- 

 netic declination, therefore it follows that the eleven-year 

 and the greater periods, both for the auroras and the ter- 

 restrial magnetism, must agree. The magnetic disturbances, 

 and probably also the secular variations of the earth's mag- 

 netism, stand also in intimate relation to the auroras. Dr. 

 Fritz concludes that the entire series of observations that he 

 has discussed justifies us by no means in forming, as yet, any 



