MAGNETIC SUllVEV OF .tAl'AN'. lOo 



From the Zcnitli distance!^ of the observed sturs, the latitude of 

 the place could be calculated by applying the u.sual corrections — reduc- 

 tion to the meridian and refraction. 



AVhen a «ufiticient Jiumber of .stars could not be observed, the lirst 

 observation of Polaris was utilized for the purpose of obtaining the 

 aziniutli to a closer ap])roximation, the newly e\;ilii;ited clock error 

 and latitude being employed, and the calculation made without the 

 use of the I'olaris a/.imuth table, 



xVt four stations (Nos. 52, 5(), 07. 71) observations of the sun 

 alone were possible. In these cases, the local time was determined 

 from altitude obser\ ations, and the azimuth from prime vertical 

 transits. 



On two occasions, at stations ((iä) and («SO), Polaris was ob- 

 scrNcd through clouds and the clock error was assumed. 



Section I V. 

 General Kesults of the Survey. 



I now pass to tlie consideration of the principal residts oblaincil. 



There were 81 stations in all. of which i){) beloni'cd to the 

 Northern Tarty, and ^2 to the Southern, the recreaticjn ground of the 

 Imperial University being a cominon station to l)oth Parties. 'I'he 

 Parties might with e(pial accuracy be termed the Eastern and West- 

 ern Parties, inasmuch as all the stations of the former lie to the east of 

 the longitude line 138° E.. tmd (with the exception of three, including 

 Tokyo) the stations of the latter lie to the west of that line. Iloughlv 

 speaking, this line separates the main island of dapan into two regions 

 markedlv ditferent as to their L-eoloiJ-ical <haracters. f'omi>ared to 

 the Western portion, the Eastern and ^s'orthcrn portion is highlv vol- 

 canic. Here, consequently, considerable magnetic disturbances are to 



