Prof. Norton on Molecular Physics. 269 



of the tropical year, n being an even number, the same region 

 of the earth's surface would at the close of each successive year 

 be, for a considerable interval of time about the autumnal equi- 

 nox, on the opposite side of the earth from the sun, and thus 

 would come to be traversed by strong currents running from N. 

 of E. to S. of W. (p. 273); also that at each successive vernal 

 equinox the same region would be on the side of the earth 

 turned towards the sun, and therefore in the most favourable 

 position for the currents already developed at the autumnal 

 equinox to be reinforced by the new currents*. The systems of 

 currents thus originating at such successive epochs would not in 

 general be coincident ; but it will be seen in the sequel (p. 278) 

 that each system should become subject to a motion of revolution 

 under the operation of the new effective currents annually deve- 

 loped, and that the annual rate of displacement should be differ- 

 ent for each system, unless their currents should be of equal 

 intensity, which would be in the highest degree improbable. 

 Now, if the shifting movements of the^different sets of currents 

 were unequal, the tendency should have been, in the lapse of 

 ages, to bring them all into coincidence, or to consolidate them 

 into one system in each hemisphere. In the light of Gauss's 

 investigations into the magnetic state of the earth, we may con- 

 clude that the earth has actually reached this period of its mag- 

 netic history. 



At the epochs for which n was an uneven number, two sys- 

 tems of currents should have been developed, one at each equi- 

 nox; and the intensity of each of these would have been much 

 less than that of the single system (the sum of the two equi- 

 noctial systems) answering to the epoch when n was an even 

 number. These separate systems of currents should therefore, 

 by reason of their secular movements, have tended to become 

 incorporated with the other more effective ones, which would 

 have been displaced more slowly. 



It will be seen, in another connexion, that the magnetic state 

 of the earth experiences certain changes from year to year in 

 response to the varying magnetic and electric condition of the 

 sun's surface. We may then conclude, from our present point 

 of view, that the existing system of magnetic currents should 

 bear the traces not only of the changes through which the mag- 

 netic condition of the earth has passed, but also of the mighty 

 changes that have passed over the face of the sun. 



From our present point of view we may discern the probable 

 link of connexion between the magnetism and the temperature 

 of the earth. In the paper already referred to (p. 267), a mathe- 



* It is here implied that the more effective currents are developed at the 

 equinoxes; in explanation of this see pp. 273, 276. 



