171: KN'OTT AXD TANAKADATE 



one case (at Sekiyama No 36) they were made at 3 a.m. Now it is 

 just at these hours that the magnetic declinarion is subject to tlie 

 slowest chano-e and has besides a vahie which can liardly ditfer more 

 than 1' of arc from the mean vahie for the day. Where a complete 

 series of liourly abservations thronghont a wliole day is oiitoftlio 

 question, observations of declination made a little before midniglit 

 will crive a fairly good approximation to the true mean value. When 

 the sun was used for finding the azimuth, it was usually before 8 a. m- 

 or after 4 p.m. On one occasion, however, (at Ebisu, N(X 33) cir- 

 cumstances compelled a transit of the sun to be taken at about noon. 



The dip and horizontal force observations were usually made 

 at early morning before eight o'clock or in the afternoon after four. 

 As there was but one tripod, it was necessary to remove the one 

 instrument so as to mdvC way for the other. It was found more 

 convenient generally to observe the dip first, and then make the 

 deflection and \'ibration experiments. If necessary a declination was 

 taken and the azimuth obtained as usual from two transits of the sun. 

 If it was a morning obser\ ati(^n and if the declination had been 

 obtained by means of Polaris the night l)ef()re, this final declination 

 was dispensed with. Ff it Avas an afternoon observation, however, 

 and the sun visible, a solar azinmth :ind magnetic declination were 

 always taken, to guard against the mischance of a cloudy night. 

 The dip and magnetometer observations were make liy lioth observers 

 alternating. T'hus the one who observed the dip during this set of 

 observations would operate with the magnetometer during the next 

 set. and rice imsd. The one who A\as not ol)serving recorded and 

 kept time. 



The sextant observations were all undertaken by Mr. Nagaoka, 

 who had carefully trained himself to the work before the Expedition 

 started. The sextant used was one of iVowning's (London), No. G29.'"). 



