318 Prof. B. Stewart and W. Dodgson. [Nov. 20, 



The results of Tables XI and XIII are exhibited in Plate 1, where 

 fig. 5 denotes the mean curve of magnetic periods, fig. 6 the curve of 

 Kew, and fig. 7 that of Prague. It will be noticed that these three 

 curves are extremely like one another. It will also be noticed that the 

 meteorological periods A, D, E, F, are well reproduced in the magnetic 

 curves, while, however, the two peaks, B and C, appear to be squeezed 

 together into a table-land. We think that this want of similarity of 

 these two peaks is probably due to certain local influences which we 

 have not been able to overcome, through the comparative paucity of 

 our observations. Leaving out of account this table-land, we may now 

 compare together the four magnetic inequalities A, D, E, F, with the 

 corresponding meteorological ones. 



This is done in Table XIV, which is precisely similar in construction 

 to Table VIII. 



The results of this table are graphically represented in Plates 2 and 3. 

 While there is a very striking likeness between the corresponding 

 magnetic inequalities, there is also a very considerable likeness between 

 the various corresponding magnetic and meteorological inequalities 

 representing periods A, D, F. There is not, however, a strong like- 

 ness between those representing E. 



18. But here it is well to bear in mind that, while we should expect 

 our method to indicate the various natural periods, whether in sun- 

 spots, meteorology, or magnetism, it does not necessarily follow that 

 the magnetic inequalities should be coincident in phase with the 

 meteorological ones. For instance, it is not necessary that the 24-day 

 magnetic inequality at Kew should coincide in phase with the 24-day 

 meteorological inequality ; for if it be true that in magnetism we have 

 at once a 24-day inequality, simultaneously affecting all stations, 

 influencing the earth as a magnet, in whatever way the influence is 

 produced, and also something more allied to the meteorological ine- 

 qualities, travelling from west to east, having likewise a 24- day period, 

 then it is quite conceivable that the superposition of these two mag- 

 netical causes should produce a period also of 24 days, but differing in 

 phase from the corresponding meteorological period. It will be noticed 

 that we do not presume to assert that the magnetical phenomena 

 are as we have just stated, but only that they may be so, and hence 

 that a coincidence in phase of the two curves, magnetical and 

 meteorological, is not a necessary consequence of the existence in 

 these two phenomena of a common inequality, having a period of 24 

 days. 



19. Owing to the present imperfect state of sun-spot recorsd, we 

 have not been able to obtain trustworthy results for the 16 years for 

 which we have discussed the other phenomena. The researches of 

 Dr. De La Hue and his colleagues have, however, given us the means 

 of exhibiting certain provisional results, extending throughout the 10 



