736 Dr. Gwilym Owen and Mr. R. Halsall on Carriers 



previously — namely, the retardation of the discharge by a 

 transverse magnetic field. This method, which is due 

 to Sir J. J. Thomson, is too well known to require full 

 description. It will be sufficient to say that if charged 

 particles starting from rest from the plane w=Q be subjected 

 simultaneously to an electric field X, and a magnetic field Y 

 at right angles to X, then the paths of the particles are 

 cycloids, and the greatest distance d they get to from their 

 plane of origin is given by 



2X m 



Jti 2 * e * 



Hence, if an insulated electrode be situated at a distance 

 greater than d from the plane # = 0, none of the particles 

 will reach it. It is obvious that the magnetic field which 

 just prevents an electron from reaching the electrode, will 

 be far too weak to produce an appreciable deflexion of any 

 charged particles of atomic dimensions that might be present. 

 Thus, by measuring the thermionic current to a distant elec- 

 trode with and without the magnetic field, information can 

 be obtained as to the relative proportions of electrons and 

 heavy ions in the discharge. 



The experimental tube used in our investigation is shown 

 in fig. 1. A is the hot wire charged generally to about 

 200 volts. The method of supporting the wire is shown 

 more clearly in fig. 2, where the disk BB is shown in plan ; 

 the object of this disk was to obtain a fairly uniform electric 

 field between the wire and the insulated electrode C. The 

 latter was connected to an electrometer by screened wires 

 and keys in the usual way. When experimenting at high 

 temperatures the electrometer was replaced by a sensitive 

 d'Arsonval galvanometer provided with suitable shunts, so 

 that currents over a wide range could be measured. D is a 

 cylinder of wire gauze connected to earth. This cylinder is 

 necessary to prevent the inside walls of the glass tube from 

 becoming charged up by the stream of deflected electrons. 

 If this precaution be omitted, the magnetic field has but 

 little retarding effect on the discharge. The distance of the 

 hot wire from the insulated plate ranged in the different 

 experiments from 6 mm. to 2 mm. At the latter distance, 

 with 240 volts on the wire (iridium), a magnetic field of 

 300 units produced an appreciable effect on the negative 

 thermionic current. [This gives 1 33xl0 7 for the value of 

 e/m for the electrons from iridium.] Generally, however, 

 magnetic fields of intensity 800 or 1000 were applied, and, 

 on occasion, fields of 2000 units. The wire w T as heated by 



