1862.] 195 



The following argument is advanced to show that the second of 

 these hypotheses is untenable. 



Let us compare together the two magnetic disturbances of August 

 and September 1859 and August 1860 ; and suppose the peaks 

 and hollows of the disturbance curves of these dates to be caused by 

 earth-currents. This would require that currents of the same name 

 should have simultaneously travelled between Margate and Ratnsgate, 

 and between Ramsgate and Ashford during the latter disturbance, 

 whereas during the former these currents should have been of 

 different names, that is to say, the one positive and the other negative. 

 According to Mr. Walker's observations however, on both these 

 occasions a current between Margate and Ramsgate was simultaneous 

 with one of the same name between Ramsgate and Ashford. 



Thus, if we adopt the second hypothesis, it would appear that 

 these lines ought to have been affected differently on these two 

 occasions, whereas by observation they were affected in the same 

 manner ; the conclusion is that this hypothesis does not represent 

 the truth. 



The author then shows that earth-currents observed simultaneously 

 with a very abrupt disturbance which commenced about 1 l h 1 7 m A.M., 

 September 1, 1859, would lead us to infer that the former are induced 

 currents due to sudden and rapid changes in the magnetism of the 

 earth. 



Referring now to the first hypothesis, which asserts that the peaks 

 and hollows represent small and rapid changes in the intensity of 

 the whole disturbing force which acts upon the magnet, it would 

 follow that these peaks and hollows should in this case comport 

 themselves with regard to the three elements of the earth's magnetism 

 in the same way as the whole disturbing force of which they repre- 

 sent the changes. Thus, if the tendency of the great body of the 

 disturbing force is to raise the curves for the three elements simul- 

 taneously, then a small peak in one element should correspond to a 

 peak, and not to a hollow, in the other two. But if, on the other hand, 

 the tendency of the disturbing force is to raise one of the curves 

 and lower the other two, then a peak in the first should correspond 

 to a hollow in the others. 



This is shown to be the case in the disturbances extending from 

 the beginning of 1858 to the end of 1860 ; and the author therefore 



