MAGNETIC H LI! VE Y OF .TAl'AN. 207 



tliîin the mean formula re(|uirt'.s. Tlii.s may have some .significance. 

 From the pretty scattered (li.stribulion of the stations we slionld liardly 

 expect to find them adjusting themselves into groups characterised 

 by similar errors. The disturbances are generally of such a local 

 character that they must almost certainly affect in different Avays the 

 magnetic elements at stations which are never very close together. 



And now expressing the co-ordinates in kilometres we find, for 

 the Total Force, 



F --- .4G407 + .0000508 9 - .0000302 X 

 from which we obtain, as in the other cases, 



u = 59° !()' .5 



r = .000059 



IV. — The Declination. 



As already stated, it is impossible to regard tlje Declination as at 

 all expressible by means of an equation involving only first powers 

 of the co-ordinates. Fortunate! v for the labours of calculation, the 

 general form of tlie isogonic lines for Japan may be taken as fairly 

 parabolic. On this assumption, then, the fifry selected observations 

 were combined by the method of least scpiares, and the result is 

 emb<jdie(l in the following formula expressing the Declination ( 6 ) in 

 terms of the co-ordinates : — 



Ô = 4° 53' .3 + (.241 9 - .109 A- - .000231 X')' 



The latitude and longitude coordinates ( ?, A,) are reterrc«! to 

 the Mean Station ( 36° 30' N, Lat., 137° 9' E. Long.) and measured 

 in minutes of arc. 



Table y. is constructed in the same fashion as Table II., 111. 

 and 1\'. 



If we take into account all the 79 observations — at two of the 

 Stations (Nos. 7 and 37) the Declination could not be observed — the 



