234 CHKTSTCHURCH TERM-DAY OBSERVATIONS. 



recorded all included between 6 p.m. and 4 a.m. Two consecutive hours during which no change 

 recorded are met with on March 1, 1902, August 1, 1902, and February 1, 11103. The longest interval 

 during which no Declination change as large as O'-l was recorded occurred on August 1, 1902, and 

 amounted to no less than 2 hours 22 minutes. On this day the Declination showed a range of only 0'' 1 

 between llh. 45m. and Kill. 3m. (i.e. between 11.15 p.m. and 3.33 a.m.). The intervals on March 1 and 

 February 1 during which no change occurred were but little shorter, being on the former occasion 

 138 minutes (from 1.30 to 3.48 a.m.), and on the latter 133 minutes (from 0.25 to 2.38 a.m.). On 

 March 1 and August 1 there were a few minutes' loss of trace during the intervals specified, but none on 

 February 1. 



In the case of H the greatest number of changes in one day took place on November 15, 1902, the day 

 for which Declination data were lacking ; the changes averaged nearly 1 9 an hour, though the absolute 

 range was only 47y. November 15 is, perhaps, the outstanding example of an almost incessant oscillation, 

 the tide of magnetic change never setting in one direction for any length of time. The immediately 

 previous term day, November 1, with exactly the same absolute range, had a number of changes less by 

 nearly 200. The smallest number of changes, 136, occurred on April 15, 1902, the same day as gave the 

 minimum number for D. The greatest number of changes of H in any one hour was 47 on March 15, 

 1903, the hour representing 4.30 to 5.30 p.m. The range encountered during this hour was only 8y. 

 There are 18 hours in Table IIlA during which no change of H as large as ly was recorded, and 

 of these seven were included between hours 14 and 16 (i.e. 1.30 to 3.30 a.m.). 



On August 1 and October 1, 1902, no change in H was recorded during two consecutive hours ; on 

 October 15, 1902, during three consecutive hours; and on May 15, 1902, during four consecutive hours. 

 The complete interval without change on this last occasion extended really to 4 hours 41 minutes (from 

 11.29 p.m. to 4.10 a.m.). Trace, it is true, was lacking for three short periods, amounting in all to 

 10 minutes, and it is, of course, possible, though unlikely, that a measurable change may thus have failed 

 to be recorded. 



In the case of the Vertical Force the number of changes (the minimum change being ly as with H) 

 showed a maximum of 61 on March 15, 1902, and a minimum of 21 on May 1, 1902. The largest number 

 of changes in one hour, 12, occurred on March 1, 1903, shortly after midnight. No change was recorded 

 during four consecutive hours on June 15, 1902, and February 15, 1903, while on December 15, 1902, no 

 change was recorded during 5-f hours (from 10.30 p.m. to 4.15 a.m.). The sensitiveness of the H and V 

 magnetographs was for H 1 cm. = 46y, for V 1 cm. = 27y. There is thus nothing in the sensitiveness to 

 account for the extraordinary excess in the number of changes in H. There is, however, some room for 

 doubt whether in the case of V absence of apparent change necessarily meant absolute uniformity in the 

 field. The discontinuities presented on several days are rather suggestive of friction between knife-edge 

 and plane, which might suffice to prevent any response on the part of the magnet to minute oscillations in 

 the field. 



If the term days are arranged according to the number of changes of reading, the order is by no means 

 the same for the different elements. This is even more strikingly true as regards individual hours. For 

 example, during hour 5-6 on March 15, 1903, whilst the changes in H numbered 47, those in D and V did 

 not exceed the average. Thus what may fairly be described as " restlessness " by no means always affects 

 the different elements simultaneously. The fluctuations may thus answer for many consecutive minutes, 

 sometimes even for hours, to disturbing forces acting persistently in one general direction. 



12. When comparing the number of changes in D and H it is necessary to remember that the minimum 

 changes, viz. /- 1 in D and ly in H, are arbitrary quantities and do not represent equal disturbances. 

 The force perpendicular to the magnetic Meridian required to produce a change of O''l in Declination 

 varies as the local value of H. For Christchurch it amounts to 0-66y. Thus the unit for changes in H 

 was really 50 per cent, larger than that for changes in D. The number of changes must decrease as the 

 size of the unit is enlarged, but the exact relationship between the two quantities is uncertain.* If we 



Taking June 1 and December 1, the effect was tried of taking 0' '2 as the unit change in D and 2y as the unit change 

 in H. The number of changes in D was reduced only about 50 per cent., but the number in H was reduced by 70 per cent. 

 The mean of these two days represented fairly average conditions so far as the total number of changes is concerned. 



