DETERMINATION OF THE GAUSSIAN MAGNETIC CONSTANTS. 



129 



Comparison of the Values of the Magnetic Constants in British Units as deter' 

 mined (1) by Gauss, (2) bij Ennan, (3) by Adams for 1842-45, (4) by 

 Adams for 1880, with their yearly rate of increase from 1845 to 1880. 



The multiplier for the conversion from British units into c.g.s. units is 0^046108. 



It will be seen on examining these tables (1) that p^J and f/°are numerically- 

 very much larger than g\, and (2) that the values of gl from the same 

 equations differ greatly according as gl is or is not included, the value of 

 i^S being —-0276 -when gl is included, and —•0124 -when gl is excluded. 

 It also appears from the comparison of the solutions -when the equations 

 of condition are included up to 77^° latitude -with the solutions above 

 {i.e., stopping at 67i° latitude) that ^2= — -0126 in the first case, and — •0276 

 in the second case, and that this arises from the fact that the sum of the 

 absolute terms in the final equation for g\ is + -08815 -when we stop at 

 latitude 67^°, and --07184 when we proceed to latitude 77^°. Hence a 

 wide variation in the value of g\ is to be expected in the two cases, even 

 when gl is included in both sets of equations. It also appears from 

 the above tables that those constants in the values of which Gauss and 

 Erman greatly differ are those which have undergone the greatest apparent 

 changes in the interval from 1845 to 1880, and that the values for 1845 

 now determined for the most part agree more nearly with those of Erman 

 than with those of Gauss.' 



The values of the magnetic constants have been determined from the 



' It should be remembered that before the excellent Admiralty charts of 1880, 

 prepared by Captain Creak, the magnetic charts of the world were based on obser- 

 vations which were insufficient and not distributed widely or regularly enough over 

 the earth's surface to lead us to expect a close agreement between the results of 

 Gauss's theory as derived from the earlier observations as compared with the later 

 more trustworthy observations. 



1898. K 



