On Rapid Periodic Variations of Terrestrial Magnetism. 35 



does not keep pace with moon's phases. It will be more plausible 

 to suppose that the period in question is related to the sun's 

 rotation. ^^ 



It is to be regretted that our observations were not very 

 adequate for investigations of the kind mentioned above, since 

 in the warmer half of the year the records were occasionally 

 defective. For the same reason, we must unfortunate^ refrain 

 from carrying out the statistical investigations with respect to the 

 annual or seasonal variation of the daily frequency. 



11. Probable relation of the frequency of waves with meteor- 

 ological elements. Since the phenomena of the magnetic pulsations 

 were suspected of being closely related to some atmospheric 

 phenomena as will be seen later on, it seemed possible that they 

 might have a sensible correlation with some of the meteorological 

 elements. Among the latter elements, cloudiness was considered 

 as a most promising one for investigation, since its correlation 

 with the solar activity is already acknowledged by some au- 

 thorities,^-' and moreover, considering the layer of cloud as a 

 conducting sheet covering a considerable part of the earth's 

 surface it may in some way or other pla}^ a sensible rôle where 

 electromagnetic waves of not indefinitely long periods are in 

 question. To carry out the comparison effectively, it will 

 however, be necessary to take the cloudiness over a sufficiently 

 wide area of the earth's surface in order to eliminate local 

 irregularities of secondary nature which are very conspicuous 

 in the case of this element. As the only available data in hand, 

 the Monthly Reports of the Central Meteorological Observatory 

 were used. Different groups of the local stations were taken 

 and the average amounts of cloudiness were calculated. The 

 results of comparison were not conclusive, though some correlation 

 was suspected between the frequency of the magnetic pulsations 

 and the average cloudiness for Tokyo, Tyosi, Mito and Maebasi, 



1) Maunder and Marchand, loc. cit. Curiously enough, Schuster, on examining 

 Maunder's data, also noticed a pronounced 14 days period. If the 14 days period was to be 

 invariably found in allied phenomena, then the seat of the direct or indirect cause of 

 magnetic disturbances may be sought in two antipodal portions of the sun's surface. 



2) Hann, Lehrbuch der Meteorologie, 2nd. Ed., p. 47ö. 



