1908] on the Carriers of Positive Electricity. 179 



tive ion at (/, this ion as it moves np to the catliode will not follow 

 the path Q' P but some such path as Q' P', striking the cathode at P', 

 and producing a cathode ray at P'. Thus the positive particle, if it 

 strikes the cathode at all, will not give rise to a cathode ray to replace 

 the one which produced it, but a ray startiug from some other region. 

 If, however, the line of force starting from P were a straight line, 

 the positive particle produced by the ray at P would strike the cathode 

 at P. AVhen the discharge is in a steady state the number of cor- 

 puscles coming from any reyion must be proportional to the number 

 of positive particles falling on that region. Now this will be the case 

 when and only (except perhaps in very special cases) when the posi- 

 tive rays which strike the region are those produced by the corpuscles 

 coming from it. For this to happen, the lines of force from that 

 region must be straight lines. In the case of the triangular cathode 

 there are six regions where the lines of force are straight, the middle 

 points of the sides and the corners ; and it is therefore from these 

 regions that we should expect the discharge to be concentrated ; and 

 inasmuch as the region over which the lines are approximately 

 straight is much greater at the middle points of the sides than at the 

 corners, we should expect the maximum discharge to come from the 

 middle point of the sides. 



Magneto-Cathodic Rays . 



Among the rays which sometimes occur near the cathode are some 

 observed by Villard,* and called by him magneto-cathodic rays. These 

 rays occur when the discharge-tube is placed in a strong magnetic 

 held ; they are in the direction of the lines of magnetic force, and 

 when subject to an external electric field they are displaced in a direc- 

 tion at right angles to the electric and also to the magnetic force. 

 Ions moving through a mediimi in which they are under the action of 

 electric and magnetic forces, and which resists their motion with a 

 force proportional to their velocity, would behave exactly in this 

 manner. For, just as a stone falling through a resisting medium 

 moves at first with nearly uniform acceleration but after a time settles 

 down into a state where the velocity is constant and equal to what is 

 known as the limiting velocity, so ions, when exposed to electric and 

 magnetic forces and to a resistance proportional to their velocity, will 

 after a time settle down to a state where the velocity is uniform. The 

 time required to reach this state is inversely proportional to the resist- 

 ance when the velocity is unity and directly proportional to the mass 

 of the ion. I have shown in my " Conduction of Electricity through 

 Gases "f that when the ions have reached this state, and the magnetic 



* Comptes Rendus, cix. p. 42, 1905. 

 t Second edition, p. 106. 



