512 



SCmNGE. 



[N. S. Vol. II. No. 42. 



or even four months in one position, and then 

 the system conformed to the same curve 

 if inverted. The explanation of this sin- 

 gular and apparently irrational phenome- 

 non was not found for a long while, although 

 the fact that it existed could not be doubted. 

 The mode or law of this inversion is yet 

 a subject of study, and its great irregu- 

 larity makes it difBcult to thoroughly un- 

 derstand. 



It was now seen that the result of mass- 

 ing all the observations on one period was 

 to give back very small residuals, which ex- 

 pressed merely the excess of the strength 

 or the energy of one system over the other. 

 Two large inverse tj'pes may thus give ex- 

 ceedingly small residuals, if they are nearly 

 equally balanced in power. The form of 

 the curve at first obtained was also par- 

 tially defective for the same reason, three 

 minor crests being suppressed in the mag- 

 netic field. On applj'ing the clue thus ob- 

 tained to the temperatures of the north- 

 western districts of the United States the 

 two types emerged unmistakably, and also 

 a curve more precisely representing the 

 normal solar field. Now on reviewing each 

 pei'iod of the European magnetic field with 

 this improved curve it was comparatively 

 simple to separate all the observed vectors 

 into two parts corresponding to the direct 

 and the inverse types. Thus the residuals 

 were greatly improved, the forms of the 

 curves steadied, and in every sense the 

 future of the problem greatly strengthened. 

 The critical elements at the earth, the 

 magnetic field and the meteorological tem- 

 peratures and pressures in the northwest, 

 all agreed in classification under the 

 same double system, and in producing 

 curves that are merely the inverse of one 

 another. 



It next became extremely important to 

 discover some simple, rational cause for 

 this peculiar inversion, so persistent for 

 years and so universal on the earth. In 



the magnetic field it is common simultan- 

 eously to all stations in the northern and 

 the southern hemispheres, in whatsoever 

 longitude. In the meteorological system 

 the principal center of concentration is in 

 the northwest of the American Continent, 

 so far as explored; other centers of action 

 will doubtless be detected. Supposing the 

 seat of inversion to be in the solar action, 

 it was proper to classify the sun-spot areas 

 in two groups, keeping the northern and 

 the southern hemispheres independent of 

 each other, by massing these respectively 

 on this period. The result is verj' gratify- 

 ing, for they give back the same funda- 

 mental curve, the southern hemisphere cor- 

 responding to the direct type, and the 

 northern to the inverse type. This fact 

 may well be regarded as the keystone to 

 our arch, as it gives stability to the en- 

 tire structure of the research. We must 

 conclude that the sun emits two types of 

 magnetic energy, whose products, to some 

 extent, are the coronal stream lines and the 

 sun-spot system, and probably other phe- 

 nomena on the sun; while on the earth is 

 to be found the same periodic function dis- 

 played in the variations of the magnetic 

 field, and the American meteorological sys- 

 tem certainly, acting continuouslj^ through 

 the curved lines of magnetic force; spas- 

 modic action within the solar nucleus gives 

 the auroral display, the magnetic storms 

 and electric currents, and possibh' other 

 important effects. 



On the sun the minor results are the 

 proof that the solar nucleus rotates with 

 very nearly the same motion as the visible 

 equator; the center of the coronal belt is in 

 latitude±55°; the sun spots drift anti-rota- 

 tionalljf, and by comparison in this period 

 all surface currents can be detected and 

 analyzed in longitude as well as in lati- 

 tude; the density of the sun must be greatly 

 modified by reason of a distribution of 

 matter into a nucleus and a distant en- 



