in different parts of the Northern Hemisphere. 145 



observation ; and although more exact observations have shewn 

 that these laws are not strictly true, in places where the dip 

 differs considerably, yet the agreement is too close to admit 

 of a doubt that they are on general principles correct, and that 

 the discrepances are due to some cause which it would be satis- 

 factory to be able to explain. Let us see, therefore, how far the 

 supposition of an effect, due to unequal temperature, may be 

 calculated to furnish such an explanation. 



In the first place, it may be proper to remark, that whatever 

 might be the temperature of the entire sphere, so that it were 

 uniform, the laws would remain the same ; the intensity would, 

 indeed, be different, but the formulae having reference only to 

 the relative intensities, dip, &c. in different latitudes, would, of 

 course, remain the same under all uniform temperatures. 



I propose to examine this question only on broad principles, 

 without attempting any thing like an analytical investigation of 

 it ; because, to investigate this problem, in all its generality, 

 would require, in the first place, a more perfect knowledge than 

 we yet possess, of the proportional magnetic developement under 

 different temperatures ; and, secondly, if this were known, it is 

 perhaps doubtful whether the modern analysis, even in its pre- 

 sent high state, would be competent to contend successfully with 

 all the conditions of such a problem. At the same time, it vtdll 

 be easy to shew, that the circumstances alluded to, viz. an in- 

 creased heat about the equator, would alter the laws which are 

 applicable to an uniform temperature, and that this change 

 would lend to results very strongly resembling the known con- 

 ditions of terrestrial magnetism. 



Let us conceive the uniform tempe- 

 ratures, to which the laws apply, to be 

 that belonging to the magnetical lati- 

 tude, in which the needle is at right 

 angles to the terrestrial magnetic axis^ 

 so that the dip is equal to the magne- 

 tic colat., then 2 tan. l=:tan. colat. 

 Z, or 2 tan. /=cot. lat., which gives 

 lat=:35°.44. 



We may then, according to the 

 broad view we are taking of the sub- 

 ject, consider the needle as placed in 

 APRIL JUNE 1827. 



