124 Astronomical and Nautical Collections. 



of them in one case as much as .00218, and in almost every 

 case much more than they differ from each other. 



After all, however, it is only in one or two of the cases 

 computed, that any of the differences amount to one thou- 

 sandth of an inch, and either of the tables may therefore be 

 considered as sufficiently accurate for the practical correction 

 of the height of the barometer. The error of Lord Charles 

 Cavendish's experiments appears, in one or two cases, to 

 have been ten times as great as this, which may have hap- 

 pened without any particular negligence in the observer. 



ii. Remarks on the Change of Inclination of the Dipping 

 Needle. ^?/Mr. Arago. Anmiaire, 182'7 . p. 207. 



At Paris the declination of the compass disappeared alto- 

 gether in 1666, and, becoming westerly instead of easterly, 

 continued to increase until 1819, when it was 22j°, and has 

 ever since that time been slowly diminishing. 



The dip has continually diminished, from the date of the 

 most ancient observations to the present time: it is now 

 about 68°. 



When the horizontal needle was found to be stationary, it 

 " was natural to suppose,^' that the dipping needle would 

 become stationary also, and that its inclination would after- 

 wards be increased. But this conjecture has not been con- 

 firmed by observation ; and the dipping needle, on the con- 

 trary, continues, as in times past, to approach nearer to the 

 horizontal position. This is a circumstance which must be 

 explained by those who attempt to investigate the cause, 

 hitherto entirely unknown, of these singular changes. 



[With all due deference to the opinion of so cautious and 

 accurate a natural philosopher as Mr. Arago, the law, if not 

 the cause i of these singular changes, may very easily be 

 accommodated to the facts in question. For the simplest 

 possible theory of the change of magnetism, is to suppose 

 that the magnetic pole performs a certain orbit round the 

 pole of the earth ; and whether regular or irregular in its 

 form, this orbit must, at some one point on each side, have 

 for itg tangent a great circle passing through the point of 



