26o SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 70 



exists here a considerable disagreement between the various curves, 

 and on this account alone there cannot be expected a very satis- 

 factory result of a comparison of these curves with the curves IV 

 and V for the air pressure difference, and for the air temperature 

 in Norway. The reader will see that at certain times the fluctua- 

 tions in these curves go oppositely to the fluctuations in the promi- 

 nence curves, and at other times in the same direction; but if one 

 imagines that there is part of the time a coincidence and at other 

 times a displacement of one or two months, he sees for example 

 that the variations in the curve for the air pressure differences for 

 the time after 1903 goes quite well with the variations in the 

 prominence curve for Catania. 



In curve III we give the monthly variations in the sun spots ; but 

 the agreement between this curve and curve IV for the air pres- 

 sure difference in the North Atlantic Ocean is also not very good. 

 Occasionally we find that the variations from month to month in 

 the sun spots go almost exactly inverted to the variations in the 

 air pressure difference, and at other times, on the contrary, we find 

 them in the same direction. It appears as if occasionally a dis- 

 placement of a month or more was brought about, after which 

 interval the variations in the air pressure difference follow the 

 fluctuations in the sun spots. This is, for example, the case if 

 one considers the great variations in the time after 1903. 



In the curves VI and VII are shown the monthly anomalies for 

 the variations of declination in Christiania and for the disturbance 

 of the magnetic elements at Potsdam. The fluctuations for longer 

 periods are eliminated because the successive twelve-monthly means 

 have been subtracted from the directly observed mean values for 

 each month.^ 



We see that these two curves present a rather fragmentary 

 agreement. Compared with the curves for the air pressure dif- 

 ference in the North Atlantic and for the temperature in Norway 

 we find here also the same conditions that were earlier remarked, 

 namely, that they run partly directly with these curves and partly 

 oppositely. It is therefore difficult to find a fixed rule in the 

 matter. We refer for further details to the curves themselves where 

 the relations are shown plainly to the eye. 



* Since the curves represent the monthly anomalies of the variations it 

 follows that the half year and whole year periods in the yearly declination 

 variations are principally eliminated. 



