230 



CHRISTCHURCII TERM-HOUR OBSERVATIONS. 



CHANGES of Christchurch Declination during Intcrniitional Term Hours, 1 902-3 (continued). 



4. To make sure that the tables will be correctly interpreted by the reader, let us consider the 

 Horizontal-Force data for the first term day, March 1, 1902. At Oh. Om., G.M.T., when the term day 

 commenced, the value of H was -22672, and no departure from this of as much as -00001 (the unit 

 adopted) was recorded until Oh. 3m., when the value had fallen to -22671. During the next minute it 

 rose again to its initial value, and then, after the lapse of 3 minutes, fell a second time to -22671. So 

 it went on, now up, now down, with on the whole an upward tendency until Ih. 15m., when the value 

 recorded was -22680. This persisted for the next 16 minutes i.e. until Ih. 31m. when a break of 

 6 minutes occurred in the observations, indicated by a dash in the time column. Readings were resumed 

 at Ih. 37m., when the force had risen by -00004. Details of what occurred during the 6-minute interval 

 are necessarily unknown. 



The term hour on March 1, 1902, was from 2h. Om. to 3h. Om., and a considerable number of the 

 changes recorded during this hour were at 20 seconds (0-3 minutes) or 40 seconds (O'G minutes) after 

 the exact minute. Observations at 20-second intervals, as has been already explained, were as a rule made 

 during several hours of each term day ; but in addition to the sheets in which these observations appeared 

 there were others which recorded only the results answering to 1 -minute observations, and except for 

 the actual term hours I have confined myself to the latter. Occasionally observations of turning-points 

 taken at fractions of a minute were inserted in the 1 -minute tables, and when these existed they were 

 made use of. 



In the case of H the first three significant figures were always either -226 or -227, and the difference 

 between two consecutive readings was never large. It was thus seldom necessary to print more than the 



