52 



SCIENCE 



[N. S. Vol. XXXIII. No. 837 



the possibility of a magnetic fluctuation 

 due to such a cause until one day he put 

 this query to me: "If a magnetic eifect is 

 produced when such a small body as the 

 moon comes between the sun and the earth, 

 why do we not have an effect every day 

 owing to solar radiation being cut off from 

 one half the earth by the other?" Pro- 

 fessor Newcomb had momentarily for- 

 gotten that we do have a daily effect of the 

 very kind he had in mind, viz., the diurnal 

 variation, and when I pointed this out to 

 him he appeared convinced as to the possi- 

 bility of a magnetic effect likewise during 

 a total solar eclipse. I have shown that 

 the magnetic effect during a total solar 

 eclipse is precisely similar to that of the 

 diurnal variation, differing from it only 

 in degree and that the amplitudes of the 

 respective oscillations are in direct propor- 

 tion to the areas of the interposing discs. 



Please recall also that in the second type 

 of disturbance above treated, viz., those of 

 December 29-31, 1908, we had further evi- 

 dence of perturbations occurring only in 

 the daylight zone; the inhabitants on the 

 other side of the earth did not experience 

 the perturbations. 



If then so much can be explained or, let 

 us say, suggested, on the supposition of an 

 existing primary electric field in the region 

 above us, then it behooves us to do our 

 utmost to find out all we can as soon as 

 possible regarding this field. And here is 

 where the great value of the extensive 

 magnetic operations of the Carnegie Insti- 

 tution of Washington will be demonstrated, 

 for the precise characteristics of that out- 

 side electric field can not be accurately de- 

 termined until the completion of a general 

 magnetic survey of the globe. Wlien that 

 has been accomplished, which we hope will 

 be the case within the next five or ten years, 

 then the constants, or rather the determin- 

 ing coefficients, can be derived for the 



various constituent portions of the earth's 

 total magnetic field. 



Before closing this section let me call to 

 mind a fact that is frequently overlooked, 

 that our only cognizance of the earth's 

 magnetic field is through its external lines 

 of force. Cut these out, and we would con- 

 clude, in accordance with our usual test, 

 that the earth was not magnetized. But we 

 might have closed magnetic systems within 

 the earth similar to that of a magnetized 

 ring. Such a ring, however strongly it 

 may be magnetized, has no outside mag- 

 netic effect and if we had no previous 

 knowledge of its internal magnetization, 

 we would conclude from our usual experi- 

 ments that it is non-magnetic. I know of 

 no method of disclosing such magnetiza- 

 tions as that of the ring without producing 

 some mechanical change in the ring itself. 

 Accordingly, our knowledge of the earth's 

 primary magnetism would be confined 

 wholly to external effects, were it not for 

 the fortunate fact of the variations con- 

 tinually caused in the earth's magnetism 

 by outside forces. You will therefore see 

 the point to my statement that, in my be- 

 lief, it is useless to attempt an explanation 

 of the origin of the earth 's magnetism until 

 ive have found o%it what causes it to vary. 

 Perhaps even then it may turn out that we 

 shall have to be content with simply rais- 

 ing the question already put by Schuster 

 whether "every rotating mass may not be 

 a magnet." 



In this connection let me record here an 

 interesting fact which I found some years 

 ago. If we determine the earth 's magnetic 

 axis and intensity of magnetization per 

 unit volume separately for various paral- 

 lels of latitude, then there is a distinct 

 connection shown in the values of the con- 

 stants involved with the speed of rotation 

 of a particle on the parallel concerned. 



