NO. 6.] INTRODUCTION. CHRONOMETERS. XLVII 





10. Increase of brightness very slow, l. 111 ^ later Sat. tolerably bright. 



11. Calculated time for D 2h 17 42". Observed continually till 2h 28m, but no de- 

 crease of brightness perceptible. Ceased for a while, but between 2 h 42"> and 

 2h 45m Sat. as bright as before after which we left the Satellite to its fate. 



12. Calculated time for D Hw 20h 6 5s. Observed til 20'' 10, unchanged, some- 

 what feebler than the other Satellites. Calculated time for R Hw 21h 16m 29* ; 

 observed 21^ ll m -20 m , but no increase of brightness. 



13. 26 s later brightness estimated as 40 s before the moment noted for D. At 24h34m 

 27 s the Sat. approached to the usual brightness. 



14. First glimpse. l m 33s later same brightness as the Satellite to the left of Ju- 

 piter [Sat. IV]. 



15. Had just moved the telescope ; as soon as it had come to rest, the Sat. was seen. 



16. Observation begun at 14h 3m. 



17. Waited for R till 12h 40m, but Sat. not visible; cirrostratus. 



Infl^lence of Temperature. 



As there were in each of the three years of the expedition periods of 

 several months without any determination of the Greenwich Time, it was 

 necessary to examine the general rate of the chronometer and its dependence 

 on temperature. For this purpose the two other chronometers must also be 

 taken into consideration. From the journal of daily comparisons the following 

 Table d was formed, containing the difference Kt Hw and Iv Htv together 

 with the daily relative rate of each. The last column (<) gives the mean 

 temperature (Centigrade) of the interval. From the curves registered by the 

 thermograph in Mr. Scott-Hansen's cabin the mean temperature for every day 

 was taken out by inspection and reduced to the chronometer-shelf by means 

 of the daily comparisons with the lower thermometer. The temperatures in 

 Table d are means for 10 days (the intervals between the comparisons are 

 in some few cases 9 or 11 days). As the thermograph was taken down 1896 

 Aug. 10 the temperature for the last 11 days is more uncertain. 



