Experiments on Positive Rays. 213 



and joins the upper one. There are sometimes several pairs 

 of these lines on a plate running up into the primary para- 

 bolas due to particles having small atomic weights. The lines 

 due to the heavier particles are often connected with lines 

 corresponding to the lower curve in fig. 4 a, while for these 

 heavy particles the upper curve is absent. 



The separation of the lines in a pair such as is represented 

 in fig. 4 a is much more pronounced when the magnetic field 

 considerably overlaps the electric, and both are fairly uniform. 

 Such a case is represented in tig. 5 (PL IX.), which is 

 ti reproduction of a photograph taken when the pressure in 

 the observation-tube, i. e. the part of the tube between the 

 cathode and the camera, was considerable. It will be 

 noticed that these pairs occur on both sides of the vertical line 

 through the origin, showing that this type of line is produced 

 by negatively as well as by positively charged particles. 

 An interesting point is well brought out by this plate : it 

 will be noticed that there is no fogging of the plate between 

 the two lines, while there is verv considerable fogging of the 

 plate below the lower line. In this case there is very little 

 fogging above the upper line, but on other plates the part of 

 the plate above the upper line is also fogged to some extent, 

 though not so much as the part below the lower line ; the 

 space between the lines is in all cases free from fogging. 



The straightness of the lines when the electric and magnetic 

 fields are uniform shows that the rays which produce any 

 one line must be moving with nearly the same velocity, the 

 fineness of the lines showing that the variations in the 

 velocity must be exceedingly small. 



Origin of the Lines. — The lines of the type of the upper 

 one in fig. 4 a are due, I think, to particles which passed out 

 from the delivery-tube through the cathode uncharged, but 

 which subsequently, by striking against a corpuscle, lost a 

 corpuscle and became positively charged before they passed 

 out of the reach of the electric and magnetic fields. The 

 lower lines in fig. 4 a are due, I think, to particles which 

 were charged when they passed through the delivery-tube in 

 the cathode into the electric and magnetic fields, but which, 

 by combining with a corpuscle, got neutralized before they 

 escaped from these fields ; the different points on the curve 

 being due to particles which have gained or lost their charge 

 at different points of their path. Let us consider the first 

 type of secondary, and suppose that, as was the case in most 

 of the experiments, the magnetic field slightly overlapped 

 the electrostatic. In the curve fig. 4 a the initial vertical 



