620 SCIENCE PROGRESS 



luminous gas ; owing to the strong ionisation taking place an 

 arc tends to set in. The principle employed by Stark was 

 to use canal rays, and apply the potential difference between 

 the pierced cathode and a subsidiary electrode behind it, that 

 is, on the side remote from the anode. If the cathode dark 

 space is adjusted to be much greater than the distance between 

 these electrodes (about 2 mm.), a potential difference corre- 

 sponding to a field up to 31,000 volts/cm. can be applied without 

 arcing. The canal rays were observed in a direction normal to 

 their path (and so normal to the electric field), to avoid the 

 Doppler effect. The hydrogen lines Hp and H y were resolved 

 for this transversal effect into five components, the three middle 

 ones polarised normal to, the two outer ones parallel to, the 

 electric field. The extreme separation obtained with a field of 

 estimated strength of 30,000 volts/cm. was nearly as great as 

 that of the two sodium D lines (under normal circumstances). 

 The helium lines were also resolved, but into different com- 

 ponents, the effect varying with the series to which the line 

 belonged. The separation of the components seems to be 

 proportional to the field. The effect is obviously of the first 

 importance, and, as the author points out, may have disturbed 

 the finer observations of the Zeeman effect, as the application 

 of the magnetic field, by diminishing the cross section of the 

 positive column, increases the longitudinal electric field, perhaps 

 sufficiently to cause the Stark effect to appear. 



J. J. Thomson has continued his work on the positive rays, 

 examining them by his well-known method of photographing 

 the trace of the rays simultaneously deflected in an electric 

 and a magnetic field on a plate placed at right angles to the 

 undeflected beam. The form of the curve obtained on the plate 



gives the ratio — for the rays causing any particular curve, and, 



further, information concerning the velocities of the particles 

 constituting the ray in question. The most recent results are 



those which concern the lines which correspond to — = 22, 



and — = 3. 1 The former line was obtained when the residual 

 e J 



gas in the tube consisted of the lighter constituents of the 



atmosphere ; a molecule of C0 2 with a double charge would 



1 -- is taken as unity for the singly charged atom of hydrogen. 



