MAGNETIC SÜKVEY OF JAPAN. 221 



of between 1 and 2 p. ni. as i« u.sualjy the case. Whether ihi.s is a 

 local peculiarity or not cannot of course be decided from the material 

 to hand ; but the otherwise unexceptional character of the o])ser\ a- 

 tion.s would lead us to regard this feature also as something that has 

 a real existence. In any case, however, it is of interest to tind that 

 satisfactory self-consistent (observations of diurnal variation c:in be 

 made during a rapid magnetic survey, and that, notwithstanding the 

 continual change of locality, the observations are sufficiently precise 

 to indicate the monthly march of the diurnal range. This is, of 

 course, quite to be expected if there is truth in the view so strongly 

 advanced by Jîalfour Stewart, and supported by Schuster's interesting 

 application of the Gaussian theory — the view that the diurnal varia- 

 tion is chiefly due to electrical movements in the upper regions of the 

 atmosphere. In such a case, local conditions should have a compa- 

 ratively small influence. 



On two occasions the Southern l\artv took hourlv observations 

 of both the Horizontal Force and Dip. The observations are given 

 in detail in the complete taljles in Appendix B. As they do not 

 seem to bring out at all distinctly the well known features of the 

 diurnal changes, they can only be regarded as materials for obtaining 

 a specially good mean for the day. They have, consequently, not 

 been shown graphically. The two sets of hourly measurements of 

 the Horizontal Force mny be however useful as an indirect means of 

 finding the temperature co-efficients of the moment of the marv^et 

 employed. The determination at Hiroshima gave the following ex- 

 pression for the magnetic moment (M) of the magnet in terms of 

 the temperature (t) in degrees centigrade : — 



M = 488.25 — .28() (t—28) + .003 ( ^— 28 )-' 



