126 



NA TURE 



\jnne lo, i8So 



little time, probably owing to the magnetism of the earth, 

 or that of the core of the induction coil not far off, the 

 movements sometimes become more regular, and slow- 

 rotation takes place. The patches of light concentrate 

 into two or three, and the green light in the bulb gets 

 more intense along two opposite lines joining the poles 

 forming two faintly outlined patches, which slowly move 



round the bulb equatorially, following each other a semi- 

 circumference apart. 



An electro-magnet placed beneath in a line with the 

 terminals (Fig. 13) converts these undecided movements 

 into one of orderly rotation, which keeps up as long as 

 the coil and magnet are at work. 



In order to compare accurately the behaviour of the 

 molecular streams at high exhaustions with that of the 

 ordinary discharge through a moderately rarefied gas, 

 another tube was taken having the upper pole an 

 aluminium wire, and the lower one a ring. Fig. 14. It 

 was only exhausted to such a point that the induction 

 spark should pass freely from one pole to the other in the 

 form of a luminous band of light, this being the form of 

 discharge usually considered most sensitive to magnetic 

 influence. This tube was also mounted over an electro- 





Class n. Class b. Class c. Class d. 



Fig. is- 



magnet, and the two sets of apparatus being actuated 

 successively with the same coil and battery, the following 

 observations were made. 



The tubes will be distinguished by the terms "high 

 vacuum" (Fig. 13) and "low vacuum" (Fig. 14). The 

 rotation produced in each tube will be recorded in the 

 direction in which it would be seen by an observer above, 



looking vertically down on the tube, his e) e being in a 

 line with the terminals and with the axis of the magnet. 

 When the rotation thus viewed is in the direction of the 

 hands of a watch, it is called direct : the opposite move- 

 ment being called reverse. To facilitate a clear apprecia- 

 tion of the actions, an outline sketch (Fig. 15) accompanies 

 each experiment. The shape of the tube shows whether 

 it is the high or low vacuum tube, and the letter d or r 

 shows the direction of rotation. 



a. Upper pole of electro-magnets norlh. 



Induction current passing through tubes so as tOi 



make the top electrode positive. 

 Rotation in the high vacuum direct. 

 Rotation in the low vacuum direct. 



b. Upper pole of magnets north. 



Top electrode of tubes negative. < 



Rotation in high vacuum direct. 

 Rotation in low vacuum reverse. 



c. Upper pole of magnets south. 

 Top electrode of tubes positive. 

 Rotation in high vacuum reverse. 

 Rotation in low vacuum reverse. 



d. Upper pole of magnet south. 

 Top electrode of tubes negative. 

 Rotation in high vacuum reverse. 

 Rotation in low vacuum direct. 



These experiments show that the law is not the same at 

 high as at low exhaustions. At high exhaustions the 

 magnet acts the same on the molecule; whether they are 

 coming to the magnet or going from it, the direction of 

 rotation being entirely governed by the magnetic pole 

 presented to them, as shown in cases a and b where the 

 north pole rotates the molecular stream in a direct sense, 

 although in one case the top electrode is positive and in 

 the other negative. Cases e and d are similar ; here the 

 magnetic pole being changed, the direction of rotation 

 changes also. The direction of rotation impressed on the 

 molecules by a magnetic pole is opposite to the direction 

 of the electric current circulating round the magnet. 



The magnetic rotations in low vacua are not only fainter 

 than in high vacua, but they depend as much on the 

 direction in which the induction spark passes through the 

 rarefied atmosphere, as upon the pole of the magnet pre- 

 sented to it. The luminous discharge connecting the 

 positive and negative electrode carries a current, and the 

 rotation is governed by the mutual action of the magnet 

 on the perfectly flexible conductor formed by the 

 discharge. 



In high vacua,'however, the law is not the same, for in 

 cases /' and d similar arrangements produce opposite 

 rotations in high and in low vacua. The deflection 

 exerted by a magnet on the molecular stream in a high 

 vacuum may be compared to the action of a strong w-ind 

 alowing across the line of fire from a mitrailleuse. The 

 deflection is independent of the to-and-fro direction of 

 the bullets, and depends entirely upon the direction of 

 the wind. 



I have already mentioned that platinum will fuse in the 

 focus of converging molecular rays projected from a 

 concave pole. If a brush of very fine iridio-platinum 

 wire, which has a much higher fusing point than platinum, 

 be used to receive the molecular bombardment, a brilliant 

 light is produced, which might perhaps be utilised. 



A piece of apparatus was constructed in which a plate 

 of German glass was held in the focus of the molecular 

 botnbardinent. The vacuum w-as so good that no hydro- 

 gen or other lines could be seen in the spectrum of the 

 emitted light. The focus was now allowed to play on the 

 glass, when the glass soon became red hot. Gas appeared 

 in the tube, and hydrogen lines now were visible in the 

 spectrum. The gas was pumped out until hydrogen dis- 

 appeared from the spectrum. It was now possible to heat 

 the glass to dull redness without hydrogen coming in the 

 tube ; but as soon as the heat approached the fusing point 



