at the Magnetic Observatory at Prague. 419 



tervals than was done at Milan, namely, at every hour from 

 5 a.m. to 10 p.m.*. This we have done uninterruptedly for 

 a twelvemonth, viz. from July 1839 to June 1840, and I have 

 now the pleasure of laying before you a summary view of the 

 principal results. 



1. The variations of the magnetic elements exhibit a de- 

 pendence on the seasons, for there appear in certain months 

 maxima and minima of which scarcely a trace is discernible 

 in the opposite months. It is therefore desirable to separate 

 the observations of the winter from those of the summer, and 

 to consider each apart. By uniting them we should risk ob- 

 taining a daily march which would be wholly imaginary, and 

 would not correspond with the phaenomena in any month of 

 the year. Thus, for example, the declination in the winter 

 months shows regularly a minimum in the later hours of the 

 evening; in summer the minimum occurs in the morning; 

 the mean of the whole year, therefore, would show two 

 minima, one in the evening, and the other in the morning, 

 whereas the occurrence of two minima on the same day is in 

 every month of the year an exception to the ordinary course. 

 1 do not consider the difference in the time of the minima in 

 summer and winter a mere displacement arising from the 

 season ; at least it is not so in the case of the inclination, in 

 which a similar difference is observable: but this point must 

 be decided by future observations continued through the 

 hours of the night. 



In the winter months, October to March, we observed, 

 At 8 a.m. Gb'ttingen mean time, the decl. == 410'77f i^L 



At 1 p.m. the max. = 427'H J^^ 



At 10 p.m. the min. = 407'84 



In the summer months, April to September, 



At 8 a.m. the min, = 398'01 9q . q 



At 1 p.m. the max. = 427'50 J 



At 10 p.m. = 405*74 



The time of the maximum does not seem to be dependent 

 on the season ; the exact time obtained from the hourly ob- 

 servations by interpolation, taking into account the second 

 differences to quarters of an hour, is O h 45 m in May, Septem- 

 ber and December; this is the earliest: in February it is at 

 l h 45 m ; and the latest occurs in July, viz. at 2 h O m . The 

 monthly means gave the daily difference greatest in August, 



* We have now extended our observations so as to include some of the 

 hours of the night. 



f* As the observations of the absolute declination have not yet been 

 made with the requisite exactness, the variations are here given in divi- 

 sions of the scale, each division being 27*2261 seconds of arc. 



2E2 



