68 



Mr. W. Ellis. Relation between Diurnal Range of 



Intervals between successive Magnetic Epochs. 



Min. 



to 



Max. 

 1_2. 



Max. 



to 



Min. 

 2-3. 



Min. 

 to 



Max. 

 34. 



Max. 



to 



Min. 

 45. 



Min. 



to 



Max. 

 56. 



Max. 



to 



Min. 

 6-7. 



Min. 



to 

 Max. 



78. 



Max. Min. 



<o to 



Min. Max. 



89. 910. 



4-95?/ 7-6Cty 4-25y 7'15y 3'30y 8'(% 5'(% 5'85?/ 4'00// 



Intei'vals between successive Sun-spot Epochs. 

 4-60?/ 7-90?/ 4-10y 7'10?/ 3'40?/ 8'40 ? y 5-(% 6'20// 



FIG. 1. Intervals between successive Sun-spot and Magnetic Epochs compared. 



The thick line shows the variation in length of the interval between successive sun- 

 spot epochs, and the thin line that between successive magnetic epochs. 1 to 2, 

 3 to 4, &c., indicate intervals .from minimum to maximum, and 2 to 3, 4 to 5, 

 &c., those from maximum to minimum. 



These numbers are represented in a graphical form in fig. 1. They 

 fun so closely together that it became necessary, iii the figure, to 

 diminish the magnetic intervals by one hour, otherwise much of the 

 magnetic curve would have fallen so near to the sun-spot curve as to 

 become obliterated thereby. The general similarity of the variations 

 in length of successive magnetic and sun-spot intervals, alternately 

 from minimum epoch to maximum epoch, and from maximum epoch 

 to minimum epoch, is thus very clearly seen. When the variation 

 of the magnetic interval is large, that of the sun-spot interval is also 

 large, and when the one is small, the other is also small. 



The mean of the five intervals from minimum epoch to maximum 

 epoch is for the magnetic effect 4*31?/, and for the sun-spot effect 

 4'18i/; the mean of the four intervals from maximum epoch to 

 minimum epoch is for the magnetic effect 7'15y, and for the sun-spot 

 effect 7'40y. Whole period for magnetic effect 1T46?/, and for sun- 

 spot effect ll-58y. 



Taking further from the table of epochs, instead of successive 

 intervals, the successive periods, 1 3, 2 4, &c., we have 





