78 REPORT— 1886. 



I cannot help thinking that on inquiry a good many difficulties which 

 have been raised will be found to disappear, and that in view of the great 

 importance of the subject the necessary funds will be forthcoming when- 

 ever a definite scheme is proposed which will lead to certain results. 



The secular variation of terrestrial magnetism will probably require a 

 different treatment. Captain Creak has' informed the Committee that a 

 large number of observations of declination, inclination, and total force 

 distributed over the world have been collected by the late Sir Frederick 

 Evans and himself, and that he has already exhibited on a globe at a 

 recent soiree of the Royal Society some leading results of the distribution 

 of the secular change for the epoch 1880. The Committee should, in 

 my opinion, collect and tabulate all available records of the secular varia- 

 tions of the three components of the magnetic force. We cannot decide 

 on the best method of reduction until the material has been collected. 



Arthur Schuster. 



VIII. liemarJis on a Provisional WorJdng Hypothesis. 

 By Professor Balfour Stewart. 



1. From various quarters there have been brought together the 

 elements of what may be termed a provisional working hypothesis with 

 respect to the main causes of the periodical variations of terrestrial 

 magnetism. 



In this hypothesis it is supposed that electric currents in the upper 

 regions of the atmosphere may be the main immediate causes of the 

 periodical non-disturbance variations of the magnet, while small bat 

 abrupt changes in the magnetism of the earth along with secondary or 

 induced currents in the earth's moist conducting strata and also (occasion- 

 ally at least) in the upper atmospheric regions, in times of auroras, called 

 forth by these changes may account for the disturbance variations of the 

 masnetic needle. It will thus at once be seen that the regular variations 

 are supposed to be mainly due to a cause above the needle, while the 

 irregular variations are supposed to be mainly due to a cause beneath 

 the needle. 



2. The electric currents in the upper atmospheric regions which 

 cause the regular variations are supposed to originate in the motion of a 

 conductor (rarefied air) across lines of magnetic force, and it is supposed 

 that such electric currents will vary in the first place according to the 

 power of the sim as exercised in producing these atmospheric motions, 

 and ill the second place according to the temperature of the moving strata, 

 it having been remarked by Professor Stokes that such strata will become 

 better conductors as their temperature inci'eases. This increase in the 

 temperature of such strata may either be due to an increase in the sun's 

 radiation of such rays as are absorbed by these strata, or to a change in 

 their constitution with respect to aqueous vapour, a substance which may 

 be presumed to possess strong absorptive power for certain rays. 



3. With regard to the solar diarnal variation, the most pi'ominenfc 

 feature is the very simple character of this variation as far as the element 

 of declination is concerned. For the average of a year, and for all but 

 high latitudes, this variation may be represented as if due to positive electric 

 currents in the upper atmospheric regions flowing during those hours when 

 the sun has most power from the equator to the poles ; that is to say, from 



