108 



observatory, side by side with those of the magnetometers, on wires 

 specially erected ; which wires, as appears from the sequel, need not 

 be of any great length ; for the Margate-Ramsgate group, three 

 miles in length, is found to be very active, and action has been found 

 on a length of 972 yards. 



Reports are then given of the behaviour of the needles during the 

 August-September storm of 1859, followed by a Table containing a 

 series of consecutive observations which the Ramsgate clerk made 

 from August 29th to September 2nd, and which show the duration 

 and changes of the currents, the general direction in which they were 

 moving, and their comparative values ; details which were wanting 

 to M. De la Rive when he wrote on this storm. 



The author adverts to the remarkable manner in which the change 

 from a current in one direction to a current in another is brought 

 about, by no drift of anything like a " circular storm," nor by any 

 kind of axial rotation ; and in contrast to the calm minutes in the 

 midst of an active period, he cites cases of activity, which are common 

 enough in periods otherwise calm. He shows also how the value of 

 the derived current varies with the size of the wire. 



Passing from this more general view, he describes the arrangement 

 made for forming a more definite opinion of the value of the currents 

 collected, and how he proposed to turn to better account the next 

 storm-days, of which some good cases followed on August 8 to 12 

 inclusive, and on Sept. 7, 18GO, which is the last storm-day that has 

 occurred. On the days in question a good series of observations 

 were made, the results of which are given in Tables which accom- 

 pany the communication. These Tables are discussed, and the values 

 of the currents of either kind are given in detail in degrees and in 

 time ; and the means are deduced. Before discussing the results, 

 the author takes occasion to call attention to the very active habit of 

 the Margate-Rainsgate line, and to the high value of the currents col- 

 lected there, although, as before stated, its length is only three miles. 



On discussing the Tables, it appears that 1 current in 20 had a 

 duration of less than half a minute. The proportion of currents 

 above and below 5 minutes' duration was as I to 2*32. Of those 

 above 20 minutes and those between 5 minutes and 20 minutes, the 

 proportion was as 1 to 3 -8. The 1 -minute currents are most in 

 number; then, in order, the 2-min., 3-min., 4-min., |-min. and 



