ABSOLUTE OBSERVATIONS AND ADJUSTMENTS. 



125 



The variation instruments were read at each hour of local mean time in the order, bifilar, 

 declinometer, balance magnetometer, at one minute before each hour, at the hour, and at one 

 minute past, until the 11th October 1882, but on and after that date the readings were made 

 at two minutes' interval, i.e., at 58m., 0., 2m., as it was found that with only one minute's interval 

 between the reading there was a certain amount of huiTy, and consequent liability to error, 

 in recording the observations. The bifilar was read at the exact second, the declinometer 

 12 seconds later, and the balance magnetometer 40 seconds after each minute, but this latter 

 instrument took more or less time to read according to the distance it was necessary to move 

 the micrometer screw to obtain a correct setting. 



On days of disturbance observations wore also made at the Gottingen hours in the same 

 manner. 



NOTES ON THE REDUCTION of the DIFFERENTIAL or VARIATION OBSERVA- 

 TIONS ; BY G-. M. WHIPPLE, B.SC, Superintendent of the KEW OBSERVATORY. 



During the period of observation at Fort Rae all the diffei'ential or variation instruments were 

 read three times at each hour, two minutes being allowed to elapse between the consecutive 

 readings, and the mean of the three readings has been accepted throughout as the true value 

 for the hour. This does not, however, obtain on term days when the tri-horary readings 

 were not made, but the actual reading at the instant of the hour was only taken. 



The observations were all entered according to local time, care being exercised on term 

 days to correct the readings for difference in time when transcribing them from the term 

 day to the ordinary observation book. 



Declination. 



The values used in the reductions are given in the following table, one scale division of the 

 declinometer being assumed to be equal to 60" of arc. {See p. 124.) 



Table 8. 



Date. 



Scale 

 divisions. 



Corresponding 

 Declination. 



i88q 



.330 

 3.30 

 330 

 330 

 330 

 330 



330 

 330 

 330 



40 20 East 

 40 19 



18 



18 



17 

 10 



5 

 4 



40 

 4c 

 40 

 40 

 40 



40 

 40 



From this table other tables were computed, giving the true values in arc of the readino-s 

 for ever}' tenth scale division from 70 to 790. 



Forms having been prepared in accordance with the model adopted by the Vienna Con- 

 ference, the mean hourly readings were converted into declination values and entered as such 

 in their respective columns, together with the corresponding movement symbols* as determiued 

 by the changes occurring in the four minutes during which the instrument was under 

 observation. 



{■ Readings rising by oscillations. 

 i „ falling „ 



\ ,, rising b}' jerks. 

 ^ ,, falling 



' \ Readings rising steadi!_v 

 1 falling 



z stationary. 



? ^Movement u.".certain. 



