_j_j_6 BIKKKI.AM). Till: M IRWKC.I A N ACROKA POLARIS FXri'DITIOX, I QO2 1903. 



In detail, however, we shall he able to find the perturbation-conditions somewhat different from 

 those we have here described as the tvpical. The forces will alwavs, as already stated, in the extended 

 negative area of precipitation, concentrate themselves about one or several storm-centres. At the same 

 time, the negative svstems that occur at the other places will more or less disappear. Frequently there 

 is a single, comparativelv verv limited, negative svstem of precipitation, while the rest of the negative 

 current-circuit has practically disappeared. This has verv often proved to be the case at about (ireemvich 

 midnight. At about this hour, we very frequently find a powerful, well-defined and comparatively very limited, 

 negative system in the north of Kurope, while at other places round the arctic /one, no negative systems 

 of precipitation are apparent, as far as we can see from our observation-material. For this reason the 

 storms that occur at this time exhibit particularly simple areas of perturbation. It is the simplest of these 

 that we have taken first, and therein' found the elementary type of negative polar storm. 



With regard to the movement of the systems, it should be observed that it is only in its main 

 features that this takes place as stated above, differences being very frequently found in the details. We 

 have, for instance, just mentioned an example of the movement of an elementary negative polar system 

 eastwards along the auroral xone, simultaneously with the development of the storm. The cause of this 

 is, we believe, in a great measure to be found in the fact that the height of the sun above the magnetic 

 equator varies. In tig. 76 we have shown a curve that, according to Stormer's calculations, gives the con- 

 nection between the height of the centre of emanation above the magnetic equatorial plane, and the 

 deflection undergone bv the rav that goes to the origin, when we consider an elementary magnet 

 situated in that point, with its axis along the /f-axis, \Ve have thought that a similar connection must 

 exist between the height of the sun above the earth's magnetic equator and the position of the various 

 storm-centres, and that when this height of the sun alters, the various perturbation-centres will be moved 

 similarly to these "distinguished" rays in the calculations. These rays, however, will move now towards 

 the east, now towards the west, according as the height of the sun changes (see fig. 76 cS: Article 71). 

 \\V should therefore also expect to find similar conditions at the storm-centres. The finding of deviations 

 from the regular moving of the perturbation-systems towards the west, is thus only a conceivable con- 

 sequence of our theory. In the first storm in Part 11 (Article 83), we have made a comparison between 

 the positions of the storm-centres observed and the calculated areas of precipitation at the various times, 

 and their movement from time to time. We think, too, that we have found in some cases verv distinct 

 analogies, although of course there will be no question of any exact agreement. 



In Chapter I of Part II, we have principally studied the occurence and development of the various 

 polar systems. In all the perturbations, we have not only again and again found the characteristic con- 

 ditions that are touched upon here, but a number of details have also appeared that are constantly found 

 in storms of most varied character. The manner in which the polar systems break in upon one another is 

 always exceedingly characteristic. We recall, for instance, the relation between the effects of the positive and 

 negative storms at Cape Thordsen in the afternoon. If we compare the afternoon storms at this station 

 on the i 5th and 2nd January, 1883, and the first November 1882 (see PI. XXY1, XXV cV XXIII), we find, 

 as proved in detail in our previous description of the storm of the 15th January, a negative storm from 

 1 2'' to i _).'', breaking in upon a positive storm of long duration. On January 2nd there is a similar 

 phenomenon from 14'' to 16'', but the positive storm is much less pronounced. On the ist November 

 also, there is a corresponding phenomenon from 13'' 30 to i6 h , but here the effects of the positive 

 storm have almost entirely disappeared. A slight indication of a similar circumstance is also met with 

 in the storm of the i st February, 1883. It would take too long, however, to go more minutely into 

 these matters here, and we will therefore only refer the reader to the description of two storms in which 

 the characteristic conditions are especially conspicuous. These are the perturbations of the 151!! January 

 and the 1 51)1 July, 1883, in which the perturbation-conditions are perhaps most easily surveyed. 



