168 HIKKl I.AXD. 'I III'. NOKWKCIAX ATKOKA 1'OI.AKls IXI'IDilloN, [902-1903. 



('hurt II: 7'inii ;" ;<>"' nnil ;" ./<>'". 



The form of the field is on the whole unaltered, except that the strength is less. 



Although it muv appear, from a glance at the curves, as if the perturbation were fairly simple, it 

 is in realitv of a rather composite character. In the district from Axeloen to 1 Jvrafjord, there; is polar 

 precipitation. Tin-re is, on the whole, a current-svslem acting as a hori/ontal current flowing almost in 

 tin- direction from Axeloen to 1 (yrafjord. The svstem should have its greatest density to the south of 

 these two stations. On account of the comparatively quiet conditions at Kaafjnrd, the powerful effect 

 at Dvrafjord and Axeloen must he due to the fact that the currents causing the perturbation must come 

 comparatively close to these stations. These currents remain in the north rather a long time with 

 varying strength, but in about tin- same position from about midnight until 5''. 



While these currents arc acting in the north, and directlv or indircctlv producing verv faint per- 

 turbations southwards in Europe, a peculiar perturbation occurs, well-defined and powerful, but of short 

 duration, and remarkable for its universal distribution. It is the more remarkable that there is noplace 

 at which it seems to be accompanied bv storms of great violence, but appears to be as powerful in lower 

 as in higher latitudes. 



We have said that the field of this perturbation resembles in its main features that of the previous 

 6th October. In addition to this, its course is on the whole the same. 1 he two perturbations are about 

 equal in duration, increase suddenly to a maximum, and then more slowly decrease to I); and their 

 strength is about equal. The only difference is that this perturbation is most powerful in North America, 

 while that of the 6th October was most powerful in Western Kurope. 



This brief storm must thus, it seems, be classed with those perturbations which we have called 

 cvclo-niedian. 



We might suppose that the held of force- in this short perturbation was produced by a descent of 

 ravs towards the earth, similar to that towards the terrella, which occasioned the appearance of one of 

 thi' areas of light that we find in fig. 68. We will examine a little more closely into the resemblance 

 of the field of force observed, to that which was to be expected according to the experiments and Stor- 

 mer's calculations. We will however draw attention to the fact that we have not yet any experiments 

 that are exactly suited to this perturbation as regards date and hour. 



At Zi-ka-wei, Dehra Dun and Tiflis, the arrows are directed westwards, answering to the condi- 

 tions near the point at the eastern end of the patch of light. Fig. 79 distinctly shows the direction of 

 the current to be as one would expect. The north-westerly direction of the arrows in Central and 

 Northern luiropc, the south-westerly at Dyrafjord, and southerly in eastern America, correspond again 

 to the rest of the path; but there is nothing answering to Axeloen. 



It is natural to look upon the whole lield of force as a composite field, imagining it to be partly 

 formed by polar precipitation round Axeloen and Dyrafjord, but also bv precipitation in lower latitudes 

 of stiffer rays, and probably chiefly conditioned bv the latter. 



We may also mention the fact that some of the polar elementary storms already described, and 

 described only as elementary, sometimes have (ields that may be regarded as the production of cyclo- 

 median storms. The best example of this will be found on Chart 11 for the 3 1 si March, 1903 (p. 122!, 

 win-re it is ol exactly the same shape as that now under discussion. 



15y assuming a composite lield such as this, we also find an explanation of the positive values of 

 /',, which occurred in the system's area of convergence, and which thus seem to be at variance with 

 the assumption ol a single polar elementary system in the auroral zone. 



\\ e have al.-.o subsequently met with a similar disagreement as regards /',, e. g. on the 26th 

 December, where we have indicated the probability that there the rays came comparatively near to the 

 earth in lower latitudes. 'Ibis had special reference to the ravs that occur in cvclo-median storms. 



