at the Magnetic Observatory at Prague. 423 



due to the altered direction of the force itself. It would 

 seem then that the dip is subject to much greater alterations 

 than has been hitherto recognized; and this is quite con- 

 ceivable, if we suppose temperature to be one of the chief 

 causes of the variation of the magnetic elements; for if the 

 daily progress of the temperature from east to west produces 

 the large diurnal variation in the declination, its annual march 

 from south to north, and vice versa, ought in like manner to 

 occasion an annual variation in the inclination. 



4, The times of oscillation of the dipping needle showed 

 during the winter, — 



s. Diff. 



0-01657 

 0-01602 

 0-02100 



0-02359 

 000832 



At 8 a.m. a maximum = 12-86978 



At 10 a.m. a minimum = 12*85321 



At 2 p.m. a second maximum = 12*86923 

 At 8 p.m. a second minimum = 12-84-823 



In summer, 

 At midnight, a maximum. 

 At 6 p.m. a minimum = 12*55078 



At 2 p.m. a second maximum = 12*57437 

 At 9 p.m. a second minimum = 12*56605 



The midnight maximum was shown by the night observa- 

 tions of May and June. 



The dependence of the hours of maximum and minimum 

 on the season could not be recognized with certainty from 

 the observations of each month separately considered ; it ap- 

 peared, however, as if those of the forenoon observed in winter 

 approached progressively nearer to noon. 



The following are the monthly means of all the observa- 

 tions made from 5 a.m. to 10 p.m. 



1839. August. 

 September. 

 October. 

 November. 

 December. 



1840. January. 

 February. 

 March. 

 April. 

 May. 

 June. 



Time of vibr. = 



12*02318 

 11*86037 

 11*78188 

 11*90803 

 12*69148 

 13-36133 

 13-40867 

 14-00672 

 13*76645 

 12*77688 

 12-39025 



From these numbers, which are not corrected for the in- 

 fluence of temperature, or for any decrease in the magnetism 

 of the bar, we cannot trace a connexion with the tempera- 

 ture, such as has been usually supposed to exist, viz. a de- 



