ABSOLUTE OBSERVATIONS AND ADJUSTMENTS. 



127 



corresponding observed temperatures ranging from about — 15° to + 25° cent. From 

 these the mean values for every change of 10° was computed, and corrected for change 

 of zero of the instrument. 



The observations as corrected by this preliminary determination of the temperature effect 

 were plotted in a curve, and irregular readings being then rejected, a new value was found. 

 In this way a final temperature correction was arbitrarily determined, and the values given 

 below adopted for the reduction of the observations to a common temperature. 



Table 9. 



Temperature. Cent. 



Corrections in scale 

 divisions. 



Temperature. Cent. 



-15 



— lo 



- 5 

 o 



+ 5 



-25 



-16 



- 8 



o 



+ 7 



+ 10° 



+ 15 

 +20 



+ 25 



Corrections in scale 

 divisions. 



+ 11 

 + 14 

 + 19 

 + 23 



The mean hourly readings having been reduced to temperature 0° by the above table, were 

 converted into absolute values by Table 3, calculated by Capt, Dawson from the Absolute 

 Observations, and additional corrections (Table 2) for change of zero being applied, the results 

 were entered for every hour in abstracts on the forms adopted by the International Polar 

 Commission, They form the tables on pp. 142 to 153 of hourly absolute values of the 

 horizontal intensity, and are accompanied by symbols giving the nature of the movements 

 at the time of observation determined as has already been described in the case of the 

 declination, p. 125. 



Similarly daily, hourly, and monthly means have been computed, and the maximum, 

 minimum, and diurnal range calculated. 



TEEM DAY OBSERVATIONS. 



The values of the horizontal intensity have been computed for every five minutes on the 

 term days already referred to, and plotted as curves. (Plates 1-23.) 

 Term hour observations of this instrument were not made. 



VERTICAL INTENSITY (LLOYD'S BALANCE MAGNETOMETER). 



The instrumental readings as recorded are those of a micrometer placed opposite the South 

 end of the magnet, and are such that one division represents a change of "00001 C.G-.S. units 

 of force, but on account of the instrumental defects already enumerated, p. 124, the last figure 

 has not been taken into account. The reductions and values are thus only given to 

 •0001 C.G.S. 



The first step in the reductions was to make a preliminary determination of the temperature 

 correction ; this was done in the same manner as for the bifilar by ascertaining the change in 

 the scale readings when temperature altered greatly, — but corresponding readings of the other 

 instruments showed a comparative absence of magnetic disturbance, — the value so found 

 was roughly calculated to be ± 6-5 divisions for ± 1° centigrade. 



B.aving constructed a table from this value the hourly readings for each day were reduced 

 to the mean temperature of the day, and the daily means for both scale readings and 

 temperature computed. 



Next, the change in readings produced by each re-adjustment of the instrument was 

 estimated both by comparison of readings before and after such re-adjustment, which values 



