202 CHRISTCHURCH TERM-DAY OI'.SKKYATIOXS. 



origin.il scheme to the whole of the quick runs, with this exception that the curves were read at 1-minute 

 intervals only during a considerable part of the term <l;iy. readings at 20-second intervals being limited to 

 part of the day, including always the true term hour. 



After the return of the German and British Expeditions the German authorities suggested that the 

 curves should all be copied on squared paper and published in this form, so that anyone could read off 

 for himself the absolute value answering to any specified instant of time. Photographic copies were thus 

 taken at Christchurch of all the quick-run curves. 



The Christchurch material for term days, due to the combined efforts of Dr. FARR and Mr. SKKY, was 

 thus of a most comprehensive and voluminous character. 



In advising as to what should be published I was guided largely by the following considerations : 

 It was obvious that to publish the tobies in full and to reproduce all the curves would occupy an amount 

 of space and entail an expenditure which it was unreasonable to expect the Royal Society to sanction 

 unless the results were likely to prove of extreme value. Now, it so happened that the term hours were 

 unusually badly adapted for securing the objects ordinarily aimed at in simultaneous observations at short 

 intervals of time. 



The scheme was presumably due in considerable measure to the success attending a similar scheme 

 proposed some years earlier by the late Dr. ESCHENHAGEN, of Potsdam Observatory. This earlier 

 scheme applied, however, to only two or three hours, and, on one of these, as it happened, there was a 

 well-marked magnetic disturbance. The simultaneous observations at the co-operating stations enabled 

 this disturbance to be followed in minute detail by Dr. A. SCHMIDT, who succeeded Dr. ESCHENHAGEN 

 at Potsdam. But of the term days observed during 1902-3 at Christchurch there was not one which 

 presented any noteworthy disturbance. This will be readily seen on consulting the tables of hourly values 

 published in " Physical Observations," pp. 160-179. Referring to the Declination results given there for 

 Christchurch, on p. 177, it will be seen that the daily range never exceeded 13'- 1, and on eight days it was 

 actually under 5'. 



For the study of a disturbance of moderate size, whose phases develop smartly, quick-run curves 

 have some marked advantages, but when magnetic changes are small and develop slowly, the gradient 

 in quick-run curves is so slight that they look uncommonly like straight lines, and really disclose less 

 to the eye than ordinary slow-run curves. This will be readily realised after inspection of the curves in 

 Plate III, which are fairly representative of term-hour conditions at Christchurch. 



Taking these facts into consideration, I decided that the publication of the tabulations made at Christ- 

 church tabulations representing an immense amount of labour on the part of Dr. FARR and Mr. SKEY 

 was likely to be more useful than reproduction of the curves. 



3. Coming now to the tabulations, it was obvious that what we may call the " raw material " in the 

 international forms consisting of the uncorrected curve readings and the corrections might be omitted 

 without seriously impairing the value of the results. The raw material is necessary if one's object is to 

 check the accuracy of the results, but it is usual to assume such checking to have been adequately 

 performed by the observers. Another abbreviation was obviously possible, which had indeed been to 

 some extent adopted in the sheets as received from Christchurch. 



The magnetic changes were usually so slow and regular that for an element to change appreciably 

 occupied several minutes, sometimes even hours. Thus in one outstanding case the Declination showed 

 no change as large as O''l, the unit adopted, between 13h. 41m. and 16h. 3m. In this instance the 

 international form provided for the repetition of the same figure even on the basis of 1-minute observa- 

 tions 142 times. Such repetitions are avoided by the method adopted here. Put briefly, the method 

 consists in recording the value of an element only at the time when a change took place. Supposing, for 

 instance, the Declination to remain at 16" 12'"0 from Ih. 30m. to Ih. :59m., but to be 16 12'- 1 at Ih. 40m., 

 it is sufficient to record the values at Ih. 30m. and Ih. 40m., it being laid down that the absence of an 

 entry signifies no change. Even as thus reduced, the material appeared to require further reduction. 

 Accordingly, selecting a few representative term days, I investigated from which of the three elements- 

 Declination, Horizontal Force, and Vertical Force results of most value were likely to be derived, 

 was at once obvious that the Vertical-Force data promised to be the least valuable. 



