PART I. ON MAGNETIC STORMS. CHAP. III. 287 



the greater storms, out of the field, and at the same time that from the other reflector enters; but the 

 latter also passes out of the field at 2i h 39, and does not return until 23 h 25. The storm is then 

 losing strength, and at 23 h 54 reaches a distinct minimum, after which it once more increases, and the 

 light from the second reflector again passes repeatedly out of the field of observation. At o 1 ' 46 it returns 

 finally, and from that time the storm abates rapidly. 



This perturbation, as we see, developes into one long storm, though with indications of the three 

 maxima that were so conspicuous at Axeleen. 



(b) A General Characterisation oj the Conditions in Southern Latitudes. 



As in most of the preceding compound storms there here appears to be a long perturbation in 

 Europe, lasting from about i8 h o m on the 26th January to 7 h o m on the 27th. During this long storm, 

 there occur some powerful intermediate storms, with a distribution of force differing from that produced 

 by the long storm. We have here three of these sharply-defined intermediate storms; and they coincide 

 on the whole in time with the three previously-described powerful storms at Axeleen. 



The conditions at Pawlowsk are to some extent different. The //-curve there on the whole shows 

 very little disturbance, there being powerful, well-defined perturbations only during the three intermediate 

 storms. In D, on the 'other hand, there are powerful perturbations from i8 h I5 m until the morning 

 of the day following. The conditions in the vertical intensity are especially interesting. The curve shows 

 a deflection of long duration and uniform direction, answering to a perturbing force directed upwards. 



Tiflis forms the transition from the conditions in Europe to those in the south and east of Asia, 

 and these in their turn to the conditions at Batavia. 



There is on the one hand a great resemblance between Tiflis and the district Kew to Pola; there 

 is the same maximum, and the course of the perturbation is on the whole the same, the only difference 

 being that the field is turned so that the conditions in the declination most resemble the //-curve at Tiflis. 

 But on the other hand, the //-curve at Tiflis shows so great a resemblance to that at Dehra Dun, for 

 instance, that it might almost be supposed that they were taken at the same place with apparatuses that 

 differed a little in sensibility. 



At Dehra Dun, Bombay and to some extent Tiflis, the horizontal intensity has on the whole a value 

 that is below the normal. On the morning of the 27th, the normal line runs for a long distance almost 

 parallel with the curve, and does not join it until about noon on that day. 



The two last maxima are fairly distinct as far south as Christchurch, one at about 23** o m , the 

 other at o h 38". These maxima, however, are not nearly so pronounced as they are farther north ; the 

 perturbation-conditions remain more constant. 



The perturbations in the western hemisphere are on the whole weaker than in the eastern, especi- 

 ally during the first part. The first maximum, which at Axeleen assumed the character of a brief, 

 powerful, well-defined storm, is distinctly noticeable though not very powerful, at Sitka; while at the 

 other stations it is almost imperceptible. 



From 22 h I5 m on the 26th, right on to 8 h on the 27th, there is unrest. We here have the same 

 two maxima as in the eastern hemisphere, namely, at about 22 h 55 and at o h 30. 



There thus occurs in southern latitudes a long perturbation in H, with a perturbing force directed 

 southwards; and to some extent the deviations in the curves are occurring simultaneously with those at 

 the polar stations. 



On glancing at the curves, we notice a no slight resemblance between those for Sitka and those 

 for Christchurch. It is true that the perturbations at Sitka are much more powerful, but the course 

 has nevertheless a great resemblance, especially noticeable in the last maximum, at about o' 1 35. This 

 is a resemblance not infrequently observed. 



