180 Professor Sir J, J. Thomson [April 10, 



force H is so large that Huq is large compared with unity, Uq being 

 the velocity acquired by an ion when unit force acts upon it, and 

 inversely proportional to the pressure, then the ions move nearly 

 along the lines of magnetic force, but have a small component in the 

 direction at right angles to both the electric and magnetic forces. 

 In these respects they resemble Villard's magneto-cathodic rays ; the 

 ions, however, would carry electric charges, while Villard could detect 

 no charge when the magneto-cathodic rays entered a Faraday cylinder. 

 It is perhaps possible that the absence of charge may have been due 

 to the rays making the gas round the Faraday cylinder so good a 

 conductor of electricity, that no appreciable charge could accumulate 

 in the cylinder. 



On the Method ly which the Retrograde Rays acquire 

 their Velocity. 



We have seen that the velocity of those positive rays which move 

 away from the cathode, and which are, when under observation, free 

 from the influence of the electric field in the tube, is practically the 

 same as that of the Canalstrahlen which have moved up to and then 

 passed through the cathode. This result is remarkable, for in the 

 latter case the action of the electric field which produces the discharge 

 in the tube would increase the velocity of a positively charged particle, 

 while in the former it would diminish it. The fact that the velocities 

 in the two cases are very much the same suggests, at first, the sus- 

 picion that the electric field may not be accountable for any consider- 

 able portion of the velocity in either case, and that perhaps the 

 particles forming these rays may, like the a particles given out by 

 radioactive substances, start with a high initial velocity, much 

 higher than they could acquire under the electric field. If we 

 refer to the table of velocities, we shall see that there is not, 

 in the different experiments, much variation in the velocity of the 

 particles ; these are always 2 X 10^ cm. /sec, and could be generated 

 by a fall through a potential difference of about 20,000 volts. We 

 must not, however, attach too much importance to the constancy of 

 the velocity, for the range of pressure over which we can make 

 accurate observations on the retrograde rays is very limited. For 

 when the pressure gets very low, and the discharge requires a high 

 potential-difference to send it through the tube, the rays are not bright 

 enough to be observed ; while if the pressure is more than a small 

 fraction of a millimetre, the rays either do not reach the screen, or 

 w^hen they do reach it are so diffuse that the phosphorescent patch is 

 not definite enough for its position to be measured with accuracy. 

 And although even rough measurements show that at these high 

 pressures the velocity of the particles is less than when the pressure is 

 low, we k,hoald not be justified without further evidence in concluding 



