MAGNETIC SURVEY OF JAPAN FOR THE EPOCH 1895.0. 21 



early in the morning, or at the noon the 8un's transit, except 

 when prevented by unfavourable weather, for which cases the 

 rate and error were interpolated from the previous and succeed- 

 ing observations. 



As the state of the weather can not usually be relied upon, 

 the first chance of making astronomical observations was always 

 taken advantage of, l^e it the sun or the stars either in altitude 

 or meridian passage. The comparatively large telescope was 

 designed for this reason, to enable us to observe large stars 

 through mists or thin clouds ; a still larger aperture, though 

 sacrificing a little definition, will be recommendable in future. 



§ 6. Diurnal Variation. 



The corrections for diurnal variations are not applied, being 

 eliminated to a large extent by taking the mean of three or more 

 sets of observations of all the magnetic elements in each of principal 

 stations (see Complete List of Observations in the Appendix). A 

 set of experiment for the determination of the horizontal in- 

 tensity and the dip was made in the morning about 8 or 9 a.m., 

 another near the middle of the day about 1 or 2 p.m. and the 

 last in the evening about 6, and often a set Wcis made at night 

 after finishing the astronomical work, when that was done under 

 favourable circumstances. Strict times of observations could 

 not be followed on account of the conditions of travelling and 

 also because of the astronomical observations, which had to )je 

 taken at ^vhatever time that was available. 



Declination was observed at as many intervals as ])ossibh', 

 so as to enable us to draw the diurnal curves (see PI. 1. to 

 PL LIX.) at each principal station. The mean value is obtained 



