336 Dr. A. Schuster. Experiments on the 



compounds is shown by the peculiar spectrum of the glow in many 

 gases. Such a compound will behave as an electrified molecule, and 

 travel towards the positive pole, the fall of potential will increase its 

 velocity, and that will again be reduced by molecular encounter, the 

 heat thus generated will possibly decompose the molecule again ; we 

 have then two ways in which the discharge can be carried, either by 

 the electrified molecules or by the decomposed atoms. If, owing to 

 some reason, the decomposing of molecules takes place at definite 

 places in the tube instead of being distributed unequally throughout 

 the tube, we have what practically becomes a secondary negative pole 

 with its dark space. There is much to be said in favour of the view 

 that the intervals between two stratifications are due to the same 

 causes which produce the dark space at the negative electrode. It is 

 needless at present to speculate on the original cause which fixes 

 the positions of these secondary poles. 



I can see nothing antagonistic to our views in the secondary 

 negative poles discovered by Goldstein to exist wherever the width of 

 the tube rapidly widens in the direction from the uegative to the 

 positive electrode. 



The Influence of the Magnet on the Electric Discharge. 



The influence of the magnet on the electric discharge is generally 

 supposed to be well understood, but a good deal yet remains to be 

 said on the matter, for although we can in each individual case trace 

 out pretty well why a certain effect happens, it is yet impossible, 

 according to the results given hitherto, to predict beforehand the 

 behaviour of a vacuum tube in the magnetic field. 



The magnetic influence on the positive part of the discharge is 

 either compared to the effect on an elastic thread carrying a current, 

 or looked on as an effect on particles projected with great velocities 

 from the positive pole. I do not think that either of these views 

 gives a correct representation of the matter. 



Let us compare for a moment the effects which are observed when 

 a magnet acts on currents passing through liquids and gases. Sup- 

 posing a liquid cell is placed equatorially in a magnetic field, strong 

 movements of the liquid are observed, yet the lines of flow of elec- 

 tricity seem to be unchanged. I consider that the first effect of the 

 magnet on a current through a gas is exactly the same ; but then 

 comes the difference. The particles of the gas which have carried the 

 discharge conduct better than those which have not, and consequently 

 the spark continues to pass through the particles which have been 

 thrown aside by the magnet ; that is to say, the current becomes 

 deflected. As the current continues it will become deflected further 

 and further, yet there will be a tendency to fall back into the original 

 line of least resistance. The final form of the discharge will there- 



