INSTRUMENTS AND RECORDS. 73 



DISCUSSION OF THE OBSERVATIONS. 



BY 



DR. C. CHREE, F.R.S. 



CHAPTER I. 



INSTRUMENTS AND RECORDS. 



1. The magnetograplis used in the Antarctic were of the Esehenhagcn pattern, constructed by 

 O. Toepfer und Sohn, of Potsdam. As full descriptions are readily accessible,* it is unnecessary to go into 

 details here. The Declination, Horizontal Force and Vertical Force are recorded on a single drum. Answering 

 to each element there is a separate base line, and there is further a record of temperature from a metallic 

 thermometer inside the box of the Vertical-Force magnet. There are thus seven traces being simultaneously 

 produced on each sheet. The photographic paper has a width of nearly 20 cms., and so long as there is 

 little magnetic disturbance and only small variations of temperature, and the sensitiveness is similar to 

 that customary in Europe, the difficulty of keeping the traces separate and all on the sheet is not very 

 serious. In the Antarctic, however, the conditions were much less favourable than is usual. The 

 magnetic elements possessed unusually large variations, both regular and irregular, whilst the changes of 

 the external atmospheric conditions were such as to cause large fluctuations of temperature even inside the 

 hut where the magnetographs were in action. The magnetographs arrived in this country from 

 Germany at so late a date that no time remained for the observer or the staff at Kew to become 

 really familiar with them, and it was unfortunately not discovered that the quartz-fibre suspensions 

 for the Horizontal-Force magnet were all so fine as to give unduly high sensitiveness. 



The sensitiveness of the Declination magnctograph is determined by the greater or less distance of 

 the magnet from the photographic paper. In the Antarctic instrument it was approximately l' - 5 per mm. 

 of ordinate. Though rather low for ordinary latitudes, this proved much too high a sensitiveness for the 

 conditions experienced. 



The Vertical-Force magnet, as actually used, was by no means too sensitive ; in fact, during most of the 

 time its sensitiveness might with advantage have been at least doubled. It possessed, however, a high 

 temperature coefficient, and this, in consequence of the large temperature changes encountered, tended to 

 make the Vertical-Force trace shift across the paper. The large variations of temperature also gave a wide 

 range to the trace from the metallic thermometer, which possessed a high sensitiveness. 



The observer had to attempt to keep on the sheet three magnetic traces and a temperature trace, all 

 four tending to shift their positions on the sheet, and two of them (the Declination and Horizontal-Force 

 traces) subject to almost incessant fluctuations, which were not at all infrequently of the order of half 

 the width of the sheet. Under these circumstances it is not surprising that traces not infrequently got 

 off the sheet, and that those from different elements tended at times to become confused. Troubles of 

 this kind were most serious near Midsummer, when the magnetic movements were largest. 



2. The source of light was an oil lamp which had to be filled daily. Any movement of the lamp 

 moved the positions of all seven spots of light on the sheet simultaneously. The Vertical-Force trace, the 

 temperature trace and the base line common to these two traces were altered together if the Vertical-Force 

 instrument was moved. These traces could also be shifted independently by means of screws. Any 

 movement of the Declination or the Horizontal-Force instrument altered the corresponding base line, and 

 also, to some extent, the trace of the corresponding magnetic element. Thus the number of ways in 

 which traces might have their position changed on the sheet was very large, while the instruments were 

 so light that a slight touch might cause movement. As a rule, mechanical disturbances can be distinguished 

 from true magnetic changes. The latter are seldom, if ever, absolutely sudden, and when magnetic 

 changes occur which appear sudden on the ordinary slow-run magnetograph trace they are seldom 

 confined to one element ; they also, of course, have no effect on the base-line traces. Thus changes due 

 to mechanical causes in the relative positions of the spots of light, though causing extra trouble to those 



' Terrestrial Magnetism,' vol. 5, 1900, p. 59 and Plate IV. 

 L 



