88 DIURNAL INEQUALITIES. 



would rise with increase of temperature, while the magnetic moment of the magnet would naturally 

 diminish ; thus, presumably, the correction required was positive when temperature was above its mean. 



At Kew, quartz suspensions used in the Watson type of magnetograph have had temperature coefficients 

 of from 3y to 6y per 1 C. But in the Antarctic the Horizontal Force was only about a third of that 

 at Kew, so that correspondingly less torsion was put into the suspension, and the presumption, accordingly, 

 is that the temperature coefficient was considerably lower than those found at Kew. The absence of a, 

 temperature correction may appreciably influence the diurnal inequality in two or three of the Midsummer 

 months. At other seasons its influence is hardly likely to be of any importance, except in the case of 

 individual hourly readings on days of large temperature change. 



In the Vertical-Force tables the hourly values are given only to the nearest lOy, but the absolute 

 maximum and minimum are given to the nearest ly. The reasons for this are as follows : Owing to 

 dislocations in both the Vertical-Force and the temperature traces there were at times somewhat large 

 uncertainties as to the relative values of the base line at different parts of the same month. It was thus 

 felt that the retention of five significant figures in hourly values represented far more than the accuracy 

 attainable. There was also the consideration that the scale was very contracted in most months, and the 

 further consideration that readings to the nearest lOy sufficed to give diurnal-inequality data, going to ly. 

 When the Vertical-Force curves were first tabulated it seemed unlikely that anything beyond diurnal 

 inequalities would be attempted. In fact, it was not until the absolute ranges in Declination and 

 Horizontal Force had been analysed that it was decided to attempt to obtain corresponding data for the 

 Vertical Force. It was obvious, however, that in most cases accuracy to nearer than lOy was obtainable in 

 the ranges because the uncertainties attending absolute values of the base line largely disappear when 

 considering differences of readings taken on the same day and accordingly it was decided to retain five 

 significant figures in the values of the daily maximum and minimum. It was often exceedingly difficult to 

 say which of several very nearly equal ordinates (allowance being made for temperature) represented the 

 true maximum or minimum. The times of occurrence were not measured very exactly. 



22. Before dealing with the data obtained for the diurnal inequalities reference must be made to one 

 source of uncertainty to which attention has already been drawn by Commander CHETWYND (" Physical 

 Observations," p. 134). Winter Quarters was a station at which there was appreciable local magnetic 

 disturbance. Observations made on the ice in McMurdo Sound, about If miles from the Magnetic Huts 

 (see Frontispiece), gave for the Horizontal Force a value of '0433 C.G.S., or only about two-thirds of that 

 found at Winter Quarters. The differences in the other elements were much less. The Declination at the 

 Ice Station was about 5 less and the Inclination If greater than at Winter Quarters. As the Ice Station 

 was in fairly deep water, the presumption is that these differences represent local disturbance at Winter 

 Quarters. To the question what the effect of such local disturbances is on diurnal inequalities one cannot 

 give a positive answer. If we suppose diurnal inequalities to be due to overhead electric currents, as 

 Dr. SCHUSTER'S mathematical calculations indicate, and as is most generally supposed, and if no sensible 

 diurnal variation is caused by solar radiation or similar cause in the local disturbing field, then the 

 presence of the local disturbance will influence only the Declination diurnal inequality, the amplitude 

 of which will vary inversely as the value of the Horizontal Force. If this view is correct, then the 

 only effect of the local disturbance at Winter Quarters would be to reduce the amplitude of all Declination 

 changes in the ratio approximately of 2 : 3. It must be admitted that there is little, if any, positive 

 evidence of the correctness of the view. There is some evidence, on the contrary, that in highly disturbed 

 regions the effect on the diurnal inequality is much more complicated. But the local disturbance at 

 Winter Quarters was not really large. An increase of 50 per cent, in the Horizontal Force is not, of 

 course, small from one point of view, but an increase of 02 C.G.S. does not mean anything very much 

 out of the way at a place where the total force exceeded 7 C.G.S. If the source of the disturbance 

 were basaltic rock close to the surface as seems most probable from the description of the station 

 there would not appear to be much reason to fear anything more than a reduction in the Declination 

 range. Under such circumstances, no doubt, direct heating by the sun would have some effect at 

 Midsummer, but it could hardly be large. Even at the surface the range of the regular diurnal inequality 

 of temperature, when largest, was only about 4 F. 



