THE INTENSITY OF TERRESTRIAL MAGNETISM. 259 



dependent on the time of day, which, it is true, are as frequently 

 intermixed with irregular ones as in the declination ; to discri- 

 minate between them with certainty will require obsen'ations 

 continued for years. If I may venture, from the very little ex- 

 perience hitherto gained, to express a supposition rather than a 

 result, the regular change seems to consist in this, — that the in- 

 tensity decreases in the hours of the forenoon, so that it attains 

 its minimum one or two hours before mid-day, and then again 

 increases. In order, however, to obtain provisionally the quan- 

 titative ratio, I have noted the position in the morning at 10, 

 and in the afternoon at 3, on thirty days in August, 1S37. The 

 result was, that on twenty-six days, the intensity was gi-eater in 

 the afternoon than in the morning, and less on only four days ; 

 the mean difference amounting to 39 parts of the scale, or some- 

 what more than the 600th part of the entire intensity. On most 

 of these days the apparatus was also noted in the morning at 9 

 o'clock ; of twenty-eight days, there were twenty-three on which 

 the intensity was still at this hour greater than an hour later, 

 and the reverse was found to be the case on five days only ; the 

 mean difference, however, amounted in this case only to Hi 

 divisions of the scale, or somewhat more than the 2000th part of 

 the entire intensity. 



Secondly, several very extensive series of observations j^rove 

 that iiTcgular, and, at times, very considerable disturbances, and 

 varying in short intervals of time, occur not less frequently in 

 the intensity than in the declination, as analogy would have 

 led us to expect. Uninterrupted series of some length have 

 been made on these occasions simultaneously with the intensity 

 apparatus, and with the magnetometer of the observatory ; 1st, 

 on the 15th July, 1837, fi'om 6 a.m. to 6 p.m.; 2ndly, in 

 the usual magnetic term of the 29th and 30th July ; and 3rdly, 

 during the extra term of the 31st August and 1st September; 

 the observations being made at eveiy five minutes. In com- 

 paring the two series, it is obseiTable, that where the declination 

 was violently disturbed, in general great disturbances also occur 

 in the intensity*. 



By the representation of the changes of the declination and 

 of the intensity in two distinct curves, as is done for the Novem- 

 ber term in the first Part, Plate XIII., we are far from obtaining 



* In a similar way, and with the same result, observations were also subse- 

 quently made with both apparatus in the term of 13th to 14th November. 



