^rACt^'ETIC survey of japax. 1 " 7 



within ihe limits of errors of observation, the place was assumed to 

 be free from local disturbance. At two stations, Haniada (No. 73) 

 and Maizuru (No. 78), thi.s dip test was made after the regular series 

 of magnetometer observations had been completed. At the former 

 station, the values differed by as much as 20', at the latter by 3'. 

 From the nature of the environment such discrepancies w^re jusr 

 what might have been expected. 



At Minabe (No. 59) we first made a series of hourly observations 

 of the declination : and this proved such an easy matter with the 

 form of declinometer used that at all subsequent stations declination 

 observations were made in sufficient number to obtain a diurnal 

 curve. Excepting when the number of distinct observations was less 

 than five, these diurnal variations are shown in diagram (Plates XTT 

 to XY). The general smoothness of the curves is a sufficient pro<jf 



of the efficiencv of the electrcHnae^netic declinometer, to l)c described 

 below. 



The dinrnal variation of the Horizontal Intensity was observed 

 twice, namely, at Hiroshima (No 61) «mi July 29th, and at Miyazaki 

 (No. 69) in September Ist. The values are all given in the tables at 

 the end of the memoir. A comparison of the two sets of observations 

 shows how very dissimilar ni'c the measured variaHons of the horizontal 

 foree in the two cases, althc^ugh the relation between the temperature 

 and magnetic moment of the bar magnet comes out very similar in the 

 two cases. From these results it is possible to obtain in the usual 

 form an expression for the Moment of the magnet in terms of the 

 temperature. 



On two occasions, at Minabe (No. 59) on Jidy 22nd, and at 

 8hioya (No 80) on October 5th. a series of observations of the dip 

 was made throuîjfhout the day. These a'ave nothino" definite (see 

 complete list in Table) although tbe variation of declination on these 

 days was as usual. 



