SPECIAL TYPE OF DISTURBANCE 191 



the probable result, or any plan existed for combining the observations. Thus the remarkable similarity 

 in the values obtained for -^ and for x from the several months owes nothing to any preconceived i<! 



Whatever may be the cause of the phenomenon, it is clear that so far as the forces in the horizontal plane 

 are concerned the second phase may be regarded, to a first approximation, as simply a relaxation of the 

 forces to which the first phase is due. On the average, D ends by being about 4' (or 8y) smaller, and II 

 by being about 3y larger than at the start. In the case of V, however, there is something more than a 

 mere relaxation, the " recovery " during the second phase being, on the average, 3 times the drop during 

 the first phase. 



It must not, of course, be forgotten that during the disturbance the ordinary diurnal changes may 

 naturally be expected to go on as usual. The hour of occurrence being variable, it is difficult to allow 

 very exactly for this. If, however, we take the mean diurnal inequalities for Midwinter as the most 

 nearly applicable, we find that during the average occurrence of the phenomenon the regular change 

 would be practically nil in D, about - 2'6y in H, and about + 0'6y in V. The subtraction of the effects 

 of these regular changes makes but little difference to the results, especially as regards the amplitudes. 

 For the angles in the case of the representative disturbance the results in Table LXXII are altered to 



f- X- 



First phase .... 77 3' 5 51' 



Second phase .... 69 31' 20 23' 



108. As to the possible cause of the special type of disturbance, the results derived from the co-operating 

 stations suggest that its seat is, mainly at least, in the southern hemisphere, but it is clearly not a purely 

 local phenomenon. The disturbances experienced at Christchurch, on the occasions for which Christchurch 

 data existed, averaged in amplitude about a fifth of those experienced at Winter Quarters. If the 

 cause is electric currents, the absence at Winter Quarters of any large vertical component during the 

 first phase suggests a nearly uniform current sheet overhead, or else underground currents having similar 

 direction and intensity over a considerable area, or a combination of the two sets of currents. 



If we take the value 75 53' for ^ during the first phase, and assume the local magnetic Meridian to be 

 152 40' E., then the direction of the hypotheticfil currents would be 



if overhead, from N.W. to S.E. (more exactly, from 48 33' to north of west), 

 if underground, from S.E. to N.W. (more exactly, from 48 33' to south of east). 



Judging by the solitary observation on the ice* on January 30, 1904, the undisturbed magnetic 

 Meridian was about 148 E., and the inclination of the hypothetical currents to this is roughly 9. The 

 small vertical component seen during the first phase might be explained by supposing that the intensity of 

 the currents varied slightly with the distance to N.E. or S.W. from Winter Quarters. 



The fact that after the horizontal movements had subsided there remained an enhanced value of the 

 Vertical Force might be explained by supposing that the current system or systems did not really die out, 

 but moved, if overhead towards the S.W., if underground towards the N.E. Or the explanation might be 

 that the currents had in reality circular paths and tended to magnetise the Earth, producing a S-pole in 

 the neighbourhood of Winter Quarters, which usually showed marked hysteresis in its disappearance. 



In our present state of knowledge it would be pretty much pure accident if one happened to hit on the 

 true explanation. But there are certain conclusions which may usefully lie drawn. 



The existence of such a phenomenon as the special type of disturbance emphasises the importance of 

 simultaneous observations not merely from ordinary observatories, but from stations much less remote 

 from the Antarctic station. If such stations existed, and were within reach of one another by wireless 

 talegraphy, and if observations on earth currents and on the transmissibility of wireless signals were 

 included in the programme, there would be a reasonable chance of a satisfactory explanation of the 

 phenomena being reached. If the magnetograms were studied as soon as available, and striking 

 phenomena noticed as they occurred, then, if they repeated themselves, it might be possible to recognise 



* ' Physical Observations," p. 141. 



