48 



SCIENCE 



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



respects. The eifect of the previous type 

 of disturbance was to superpose, on the 

 earth's existing magnetic field, a subordi- 

 nate magnetic system possessing essentially 

 the same characteristics as the primary 

 field, differing from it only in degree — in 

 brief it increased momentarily the earth's 

 magnetization, and hence might be termed 

 a "positive magnetic perturbation." It 

 was also a world-field — its effects were re- 

 corded all over the earth. 



But now we are to have examples of a 

 "negative magnetic perturbation," whose 

 effect Ls to superpose a magnetic field oppo- 

 site to that of the earth, in short, diminish 

 the earth's prevalent magnetization, and 

 whose area of action is a comparatively re- 

 stricted one. These perturbations were not 

 felt over the whole earth within a few 

 minutes of the same absolute time; instead 

 the intervals between the recurrences in 

 different parts of the earth were to be 

 measured by hours and even a day. 



Attention was first called to these pecul- 

 iar magnetic disturbances by D. L. 

 Hazard, of the United States Coast and 

 Geodetic Survey, and recently E. L. Paris, 

 of the same organization, has collected in- 

 formation regarding them from a large 

 niunber of observatories over the globe. 

 These disturbances did not extend much 

 over one half hour and occurred on a 

 practically undisturbed day. The maxi- 

 mum deflection in the horizontal compo- 

 nent of the earth's magnetic intensity 

 amounted to about one five-hundredth 

 part. The complete data will be found 

 given by Mr. Faris in the March, 1911, 

 issue of Terrestrial Magnetism and Atmos- 

 pheric Electricity. I am indebted to him 

 for the privilege of making use of his data 

 in advance of publication. 



Pour times out of eight cases the region 

 over which the disturbance prevailed was 

 the American continent and the Pacific 



Ocean as far as Honolulu; in one case the 

 region was limited to the Atlantic Ocean 

 and the American continent; twice it oc- 

 cured in eastern Asia and but once in 

 Europe. Hence, had we been obliged to 

 rely solely upon the magnetic records from 

 the region of the earth, Europe, where the 

 majority of magnetic observatories exist 

 (about 20), we should have had to report 

 but one magnetic disturbance between De- 

 cember 29 and 31, 1908, instead of actu- 

 ally eight. No fact known to me illustrates 

 more convincingly than this the folly of in- 

 creasing greatly the number of magnetic 

 institutions in the same region of the 

 globe. It also proves that, when dealing 

 with general terrestrial magnetic phenom- 

 ena, no such weight can be attached to the 

 combined testimony of the European ob- 

 servatories as has heretofore been the cus- 

 tom. To give weight, as is frequently 

 done, according to the number of existing 

 magnetic institutions in any one region 

 leads to totally erroneous results. "We are 

 reminded of the wise words of Joseph 

 Henry which, though uttered in a differ- 

 ent connection, apply with special force 

 here, viz., "Votes in science shoidd not ie 

 counted, but weighed." 



Now why is it that these particular mag- 

 netic perturbations were confined each 

 time to but a portion of the globe? The 

 intervals in time between the successive 

 occurrences range from l"* 5" to 24'' 21™, 

 whereas the apparent velocities shown over 

 the area covered at any particular appear- 

 ance of the disturbance is on the order of 

 the quantities as previously found for the 

 first type of disturbance considered. The 

 question immediately arises, therefore, as 

 to whether we are dealing here with tivo 

 velocities. Have we, for example, a vortex 

 consisting of very rapidly moving elec- 

 trical charges, an earth-spot, as it were, 

 the vortex as a whole, however, moving 



