PART. II. POLAR MAGNETIC PHENOMENA AND TERRELLA EXPERIMENTS. CHAP. IV. 599 



he pole of the terrella with 20 amperes magnetisation the H has been approximately equal to 4100. 

 f we then calculate the stiffness of the corpuscle-rays on the earth in the same manner as above, we 

 ;hall see below that we are finding very nearly the same values as before. 



The new terrella was first placed in our former smaller vacuum-tube (fig. 200), but it appeared that 

 he cathode rays were thrown in such numbers against the walls of the tube, that hardly any reached the 

 errella, and the above-described large vacuum-box was therefore, with much labour, constructed, and the 

 errella placed in it as shown in fig. 219. A number of test experiments were also made with these 

 ipparatuses, but unfortunately no photographs were taken except the one here reproduced, in which 

 he magnetising current was 20 amperes and the tension about 2000 volts. This shows exceedingly well 

 low the polar ring approximates the proper dimensions as compared with the conditions on the earth, 

 he angular diameter of the ring being here 49; and with a magnetising current of 30 amperes we 

 ibtained a polar ring with about the same angular diameter judging by the eye as the auroral 

 :one on the earth, i.e. rather less than 45. There is, however, no photograph of this magnetisation. A 

 ew days later, a leak appeared in the vacuum-box, which a couple of months' work failed to stop. In 

 :ase anyone should hereafter like to construct such a large vacuum-box, I would advise the use of glass 

 sheets of 25 mm. thickness and not as here 22 mm. as the enormous pressure is liable to bend thinner 

 elates too much. 



There are two important conclusions that we can draw from the polar light-ring here photographed. 

 "irstly, we can by this experiment control our earlier calculation of Hq for the cosmic corpuscle-rays 

 iround the earth. If we then by a very little extrapolation calculate the stiffness of the corpuscle-rays 

 m the earth corresponding to a circle with a diameter of 45 in the same manner as above, we find that 

 '/o = 3.1 X 10", or exactly the same value as before. The second important question we can now 

 solve is that of the breadth of the band of precipitation on earth of the rays which occasion the 

 >olar magnetic storms. For various reasons I have hitherto assumed (^ that the width of this zone of 

 precipitation between Kaafjord and Jan Mayen is less than 500 kilometres. The measurement of the 

 vidth here on the night side of our terrella gives for these somewhat stiff rays that the breadth is 

 2.5 which corresponds to 280 kilometres on the earth. 



The photograph reproduced shows an experiment (pressure o.oi mm.) in which the south pole of 

 he terrella is turned directly towards the observer. The two horns of light that are drawn in towards 

 he polar regions of the terrella are here seen coincident with one another. In the photograph in fig. 

 200 the poles were above and below, and these two in-drawn horns of light were separate. 



In this photograph we also see the exceedingly interesting manner in which the greater number 

 jf the rays are thrown in a direction away from the terrella on the morning side. It is this collection 

 if rays which, in my opinion, plays an important part in occasioning the zodiacal light seen in the 

 norning. In our photograph, on the other hand, the rays that cross one another in what we have 

 :alled the first and second lines of intersection, or lines of precipitation (see figs. 201 207 and 214 

 i 215), are not visible. I think we should easily get the regions about these two lines of intersection 

 he first by preference self-luminating in the vacuum-tube, if we so arrange it that the rays that go 

 ound the terrella on the evening side are sufficiently intense. This can be attained either by bending 

 he cathode slightly upwards, so that several of the rays pass above the terrella, or by the equally 

 simple method of making the cathode exceedingly large, almost as large as the vacuum-box permits, 

 n the latter case, the conditions will be as nearly as possible like those between the earth and the 

 sun, as the pencil of parallel rays will be the largest possible. 



As will appear later on, I consider the first line of intersection (line of precipitation) of the rays 

 3n the afternoon side to be of importance in connection with the zodiacal light visible in the evening, 



I 1 ) Expedition Norvegienne de 18991900, p. a6, a. 



