393 



I have found, that the dip in Paris, obsen'ed between 1798 and 1836, may 

 be represented by the following formula : — 



t = 69° 38'.9 — ;4'.4C5 (t — 1800) + O'.02339o (t — 1800)», 



which gives the minimum t = 66" 5'.8 when t = 1895. 



That the dip cannot decrease to zero, and, accordingly, should have a mini- 

 mum, is quite probable ; by these two series of observations, this minimum 

 is likely to arrive before the end of this century, perhaps earlier in northern 

 than in southern latitudes. 



I am not quite of yoar opinion, that the observations with a dipping needle 

 should be rejected only while it gives a diflerence of a dcgi'ee and a half be- 

 tween the different positions. This difference may be derived from two dif- 

 ferent causes, — the eccentrical position of the centre of gravity, and the devia- 

 tion of the/o/-OT of the pivots from a true cylinder. In the first case the mean 

 will approximate very nearly to the true dip, if the moment of the needle 

 before and after the reversion of the poles is not too different ; and when 

 the needle is magnetized with the same four magnets, and the same num- 

 ber of strokes, this will scarcely arrive. In the second case, the needle will 

 give different errors by different dips. This error can be detected by turning 

 the axis in the needle in different observations in the same place, or by ap- 

 plying an eccentrical weight upon the axis, if it is fixed. When the mo- 

 ment of this weight is varied in the different observations, the needle will 

 repose upon different points of the pivots in different obsenations, and 

 thereby the error v/ill be diminished, if not quite eliminated. I think this 

 has been the case with your needle A.l, as the difference A.2 — A.l changes 

 its sign by diminished dip. 



■ It is also my opinion, that even the best needle may, by a single observa- 

 B tion, give an error of three to five minutes (perhaps more), including error 

 B of observation, and irregular variation of the direction of magnetical force 

 B of the earth. I shall liere offer some observations with Gambey's instru- 

 ct ment here in Christiania, in 1839. 



