l82 



NA TURE 



[December 23, 1897 



ing to two are here reproduced (F^ig4), and of course, if the 

 basis of our argument were unimpeachable, they should be 

 identical. This is not so ; there are noticeable, although not 

 very great, differences, but, together with other facts already 

 mentioned, the variations are sufficient to enable us to affirm 

 that the secular change cannot be deduced in this simple fashion 

 from the present magnetic state of the earth. 



There is one other point with regard to the curves which 

 shows how great must be our caution in arguing from such data. 

 In the lower parts there is a general similarity with the curves 

 deduced from past observations with the magnet at different 

 places near the prime meridian, but in the upper portions, 

 which refer to the future, the curves are more complicated, and 

 they are varied by irregularities and loops of which we have, as 

 yet, had no actual experience. Dr. Bauer has observed most 

 praiseworthy caution in deducing any definite conclusions from 

 these interesting speculations. But even if we refuse to accept 

 his hypothesis that there is a more than accidental connection 

 between the secular curve which the pole of the magnet de- 

 scribes at any one place, and the curve which it would describe 

 if carried round the earth in the latitude of that place, there can 

 be no question that it is possible that irregularities similar to 

 those seen in the one set of curves may in the future occur in 

 the others, and that any deduction which we may make as to the 



for any hint as to a possible physical explanation of the secular 

 change. 



In addition to the long-protracted changes which have been 

 discussed, the pole of a freely-suspended magnet also describes 

 an orbit from day to day, which is so small, when compared 

 with the secular path, that hitherto I have neglected it. This 

 diurnal change has long been studied. Its magnitude varies 

 from time to time and has been found to be connected with the 

 season of the year and with the sunspot period. On these 

 comparatively well-known facts I will not dwell, but there is 

 one point in connection with the diurnal variation which has 

 recently been attracting attention. For days together the orbits 

 described may be very similar, but this regular motion is fre- 

 quently disturbed by violent perturbations, and the pole of the 

 magnet moves within a few minutes both to the west and to the 

 east to an extent far exceeding that due to the ordinary- diurnal 

 motion, so that the form of the path is widely irregular. Such 

 a phenomenon is called a magnetic storm, and from the fact 

 that such storms occur simultaneously at places which are widely 

 distant it has been thought that it is possible that they are due 

 to some impulse which reaches the earth from the sun. They 

 are also closely connected with displays of the Aurora Borealis. 

 These irregularities make it difficult to determine what is 

 the true normal behaviour of a freely suspended magnet. The 



w ao 18 le 1* la w s 6 ♦ a 



o*'a ♦ 8 » wt 



1780 

 18S5 



LATITUDE 40° N. 



LATITUDE 40 S. 



N28 



EQUATOR. 



Fig. 4. 



period of the magnetic cycle from the past, may be woefully 

 falsified by future events. 



It is instructive to compare with this speculation the results 

 obtained by Captain Creak from a comparison of the maps of the 

 magnetic state of the earth drawn by General Sabine and those 

 which Creak himself elaborated with the aid of the observations 

 made by the Challenger expedition. The conclusion to which 

 he came was that during the forty years which elapsed between 

 the epochs for which the maps were drawn, the secular change 

 might be best represented by supposing that the magnetic poles 

 were stationary, but that at certain points on the surface of the 

 earth the forces of terrestrial magnetism had been increasing or 

 diminishing. I have before now referred to these interesting 

 observations in public, and I will not dwell on them again, ex- 

 cept to remark it is possible that we may have to deal with 

 causes both of the one type and of the other, that within the 

 earth there may be a rotating magnetic system, and that purely 

 local causes may also alter the intensity of the magnetic forces 

 at different places. 



Dr. Schuster has recently shown that a motion of the magnetic 

 pole might be produced if the space round the earth were con- 

 ducting. The conditions under which this would occur cannot 

 be discussed now, but the magnetician has reason to be grateful 



NO. 1469, VOL. 57] 



most obvious plan for determining the average movement of 

 the needle is to take the mean of its positions at each one of the 

 twenty-four hours for every day in the year. The disturbances 

 just referred to will thus be included in the calculations, but as 

 very great magnetic storms are comparatively rare, the final value 

 will not be very largely affected. This process is very laborious, 

 and exceeds the powers of any except the best equipped ob- 

 servatories. Some time ago, therefore, it was suggested that the 

 calculations should be shortened by the selection of quiet days 

 only, on which it was to be supposed that the behaviour of the 

 needle was normal. Five days in each month were considered 

 sufficient, and by a happy arrangement all the English observa- 

 tories have agreed that these days shall be chosen by the Astro- 

 nomer Royal, and that thus the determination of the normal 

 movement of the needle shall be made by all of them from data 

 collected at the same time. 



Attention has recently been drawn to the fact that, whatever 

 the convenience of this five-day method may be, it leads to the 

 conclusion that at the end of a quiet day, the needle does not 

 return to the position which it occupied at the beginning. This 

 point has been carefully investigated by Dr. Chree, and may be 

 illustrated by means of the curve which I have drawn from the 

 Kew Observatory Report for the summer of 1895, the scale of 



