AT WINTER QUARTEKS. 1(55 



This answers to 10.4 to 10.10 a.m., L.T., on the 2Gth in the Antarctic. As Plate XXXIII shows, there 

 were also very rapid oscillatory movements in the Antarctic during this time ; but there is again some 

 uncertainty about the exact identification of corresponding movements. Conspicuously the most rapid 

 H movement in the Antarctic is the one of which the turning-point is lost owing to the trace going 

 beyond the limits of registration. This consists of a to-and-fro movement, first up, then down the 

 sheet. When the rapid upward movement began the curve was moving up the sheet, but at a 

 comparatively leisurely rate. The time of the rapid acceleration of movement is almost, if not 

 exactly, 10.4 a.m., L.T. Similarly the movement down the sheet continued for some time after a 

 marked reduction appeared in the velocity. The reduction of velocity set in very approximately at 

 10.10 a.m., L.T. Thus, presumably, the very rapid part of this to-and-fro movement in II representing 

 a decrease of the element followed by a larger increase answered to the commencing movement seen 

 elsewhere. At the time when the H trace accelerated its movement up the sheet, the D trace began to 

 move down ; but after a minute or two the motion downwards diminished, and a small oscillatory move- 

 ment intervened of which details are not clearly shown in the original. Thus during the latter part of the 

 very rapid oscillation in H the value of D was nearly stationary. The V trace went through several 

 oscillatory movements. The one which probably corresponds to the double movement in II is that of 

 which the end is represented by a crest which appears nearly 1 mm. to the right of the rapid downward 

 movement in H. The previous turning-point (a hollow) comes between the upward and downward H 

 movements, and answered probably to the turning-point in H which was beyond the limits of registration. 

 The previous turning-point on the V curve is not shown, because the curve was so near the D base line 

 that it got eclipsed by the hour stop, which came on at about 10.3 a.m., L.T. The V double movement is 

 somewhat larger than one which immediately preceded it, but is in no way outstanding. 



The most conspicuous movements in Plate XXXIV are those appearing between 6 and 9 p.m., L.T. In 

 particular, there is a bold movement of the D and H curves up the sheet, commencing at 6.45 p.m., L.T. 

 Both curves went beyond the limit of registration, but the turning-point in each was apparently a few 

 minutes after 7.0 p.m., L.T. The return movements down the sheet were equally conspicuous and ended 

 about 7.35 p.m., L.T. This double movement was in all probability of the " special type," but owing to 

 the absence of the V curve our information on the point is incomplete. This movement corresponds in 

 time to that shown by the Christchurch D and V curves in Plate XIX, culminating about 8 a.m., G.M.T., 

 on the 26th. This double movement in the Antarctic was immediately followed by another lasting about 

 50 minutes, in which, however, the motion of the H magnet was opposite in phase to that of the D magnet. 

 There were large subsidiary oscillations in the H trace. Subsequent to these two large movements there 

 were some rapid oscillations of considerable size during which the D and H traces are a little difficult to 

 distinguish. 



75. Plate XXXV shows the first part of a highly disturbed Antarctic record which commenced a little 

 before 6 p.m., L.T., on October 12, 1903, and ended between 5 and 6 p.m. on October 13. The portion 

 reproduced ends at 4 a.m., L.T., on the 13th. The H trace was beyond the limits of registration con- 

 tinuously from 8 p.m. on the 12th to 8 a.m. on the 13th, whilst the D trace was off the sheet continuously 

 from 4 a.m. to 11.50 a.m. on the 13th. Subsequent to 8 a.m. on the 13th, the H trace came on the sheet at 

 the top, executed a number of oscillations proceeding gradually down the sheet, and disappeared a little 

 after 11 a.m. Subsequent to that, except for a few fugitive appearances, it was not seen until between 

 3 and 4 p.m. on the 13th. After coming on the bottom of the sheet about 11.55 a.m. on the 13th, the D 

 trace remained on, showing a number of rapid oscillations, until a little after 4 p.m., when it went off the 

 sheet at the top. After remaining off for 10 or 15 minutes it came on again and the disturbance became 

 less. 



The range in D shown in Plate XXXV is 4 36', and that in V 157y, the latter being practically unaffected 

 by temperature. Between 11.55 a.m. and 4.15 p.m., L.T., on the 13th the range in D exceeded 4 51'. 

 The V trace throughout the whole time covered by the original sheet showed numerous short-period 

 oscillations, but none of the subsequent movements were so large as that shown in Plate XXXV from 

 9.40 to 10.40 p.m., L.T. 



There was a very considerable magnetic disturbance at Kew on October 12-13, but the principal 



