PART II. POLAR MAGNETIC PHENOMENA AND TERRELLA EXPERIMENTS. CHAP. IV. 585 



in both cases was about 2500 volts. The photographs show the important fact that the more highly the 

 terrella is magnetised, the farther does the positive precipitation reach towards the evening side. There is 

 no precipitation on branch 3 with 10 amperes' magnetisation of the terrella, but it is there with 20 amperes. 



With conditions corresponding to those on the earth, where the spherical diameter of the auroral 

 ring may be put at about 45 , the positive precipitation might reach far on into the evening side of the 

 terrella. 



Nos. 1316 are from experiments in which two additional small screens were introduced. One of 

 these was square, and placed at right angles to branch 3. It was pierced with a hole, and extended 

 1.4 cm. on each side of the branch. The other small screen was also square, was furnished with a foot, 

 and placed radially in relation to branch 5. 



The purpose of these small screens was to find out whether the rays forming the positive precipi- 

 tation on the branches are only such as come by way of the poles (see the experiment in fig. 204, 

 Nos. 1315 and Art. 112), or whether that precipitation is due to other rays belonging to the system 

 of rays that first intersect the magnetic equatorial plane several times. 



In Nos. 13 and 14, the positions are similar to those in Nos. i and 3. The conditions are very 

 much the same, with from 2800 to 3000 volts between the electrodes, and about 25 amperes to the 

 terrella-magnet. 



In Nos. 15 & 1 6, the position is the same as in No. 2. The magnetising current to the terrella is 

 lo and 20 amperes respectively, with 22 milliamperes at 3000 volts in the discharge. The absence of 

 positive precipitation on branch 3 in No. 15 will be understood on comparing that photograph with 

 Nos. ii and 12. 



Some experiments were made without photographing, the magnetising of the terrella being changed 

 from 5 to 15 amperes. It then appeared that the little screen at the pole was illuminated from the right 

 when the magnetising current was 5 amperes, the light gradually moving nearer to the pole as the 

 magnetising was increased to 6, 8, 9, 10, &c. amperes. On branch 3, positive precipitation first appeared 

 with about 12 amperes, and when the magnetising current was weakened, moved out from the extremity 

 of the branch on to the left flap of the small additional screen, and finally disappeared. 



It will be seen that these experiments did not throw much light upon this circumstance; but we 

 shall now see how the facts of the case stand. 



Fig. 214 shows eighteen photographs and fig. 215 sixteen photographs of a series of experiments 

 made with this object in view. 



Nos. i, 2, and 3, fig. 214, are from experiments made with a larger screen attached at right angles 

 to branch 3, the positions being similar to those in Nos. i, 2 and 3 of fig. 213. The magnetising current 

 to the terrella was about 25 amperes, and the tension in the discharge about 3000 volts. 



It will be noticed, in No. i, how the light falls upon the upper side of the new screen, with its 

 lower edge more or less sharply defined. It should also be observed that the shadow of the suspending 

 wire, visible in the polar light-ring shows that the rays that come into the ring seem to have passed 

 above the new screen, that is to say at some considerable distance from the terrella's equator. 



In No. 2 we first notice that the positive precipitation on branch 3 is not affected in any special 

 degree by the new screen. On the other hand, it will be seen that part of the negative polar ring is 

 lost behind the screen, showing that on that side the screen has been high enough to intercept some of 

 the rays that would have helped to form the polar ring. 



No. 3 shows the same shadow in the negative precipitation, and also a peculiar light-effect to the 

 right of branch 3, and on the terrella behind the screen. This may perhaps be foreign light produced 

 by a discharge at a point on the terrella itself, a discharge that was found out during these experiments, 

 and was the occasion of their being broken off. 



