— 10 — 



At Kagoshimii n few observations were inatle in tlio Kenclio building;, 

 which from tho ahakineaa of tho floor proved unsuitable. Accordinj^ly n 

 stone block was erected on tho outside of tho building witli n sliolter to protect 

 it from tho sun's rays. Tho timo of vibration observed tliero diffori'd from that 

 observed in tlio building. Observations were then nmdu ut two oilier places ; 

 and again the results were discordant. Hut before tho cause of tlieso discre- 

 pancies could be investigated tho party were ou the point of leaving Kugo- 

 shima. The results given below are deduced from iho niean of all the observa- 

 tions made outside of tho building. 



At Naha, tho magnet 13 was accidentally exposed to tiio sun's direct rays 

 for a sliort time. Tho temperature, as indicated by tho thermomeler iu the 

 glass case rose to about 50"C, and produced a great change in tho magnetic 

 moment. Therefore as regards magnet B tho intensity observed at Kngoshima 

 is compared with tho previous determination at Tokio, while the intousily 

 observed at Naha is compared with the subsequent determination at Tokio. 

 Tho results of magnet A are more trustworthy. 



The means of tho times of vibration are found to bo : — 



At Tokio before tho excursion 



Magnet A l:5.li|:Js 



H 10.7U7U 



At Kngoshima 



Magnet A 1;J.2:331 



13 10.3005 



At Naha 



Magnet A 12.78.^7 



IJ lO.iOlt; 



At Tokio after the excursion 



Magnet A 1 ■■5.587 !■ 



15 11.81:31 



The mean of tho two observations at Tokio 

 Magnet A l.'i.r.dicj. 



The intensity being inversely proportional to tho square of time of vilira- 

 tion, the ratios of the horizontal intensities at Kngoshima and Naha to that at 

 Tokio are as follows :■ — 



Kngoshima. 

 Af , * /i.i.r,on:3\« , „.,,., 



^f-''R°^t^ \T^ni^i)=^-^''^-^ 



B (1^^=1.0000. 



VlO.:jü05/ 



