TRANSACTIONS OF SECTION A. 523 



for the dark space next the cathode to attain large dimensions before reaching 

 the walls of the tube, the pressure may be reduced much lower without risk of 

 sparking through the tube than is possible with small tubes, and hence many 

 phases of the discharge can be investigated which either do not exist or aro 

 inconspicuous when the tubes are small and the pressure necessarily high. The 

 author has used tubes with volumes as large as eleven litres, but tubes made of 

 two-litre flasks such as are used for boiling-point determinations are large enough, 

 if the anode is suitably placed, to show the various types of positive rays. 



With these tubes the following types of rays, passing through a hole' in the 

 cathode, can be made out : — 



1. Rays which are not deflected either by magnetic or electric forces. 



2. Secondary rays produced by rays of type (1). These are deflected by 

 electric and magnetic forces ; they have a constant velocity of about 2 x 10* 

 cms. per sec, which does not change, however the pressure in the tube or the 

 potential difference between the electrodes may be altered. The value of e/m 

 for these rays has the constant value 10*. 



These secondary positive rays are accompanied bv negatively changed ones 

 which have the same velocity and the same numerical value of e/m as the 

 positive ones. 



In small tubes the only rays which are prominent are those of types (1) 

 and (2). 



3. In addition to the rays of the preceding types, there are rays which are 

 characteristic of the gases in the tube. These are only conspicuous when the 

 pressure is low. The velocity of rays of this type, unlike that of the preceding 

 type, depends upon the potential difference between the electrodes. When there 

 are several gases in the tubes— say, hydrogen, air, helium— the maximum kinetic 

 energy of the rays corresponding to each of these gases is the same, and seems 

 to be that due to a fall through the potential difference between the negative 

 of low and the cathode. The value of e/m for the rays from different gases is 

 inversely proportioned to the atomic weight of the gas from which the rays 

 are derived. Thus these rays are probably atoms of the gas carrying one unit 

 of positive charge ; in the case of hydrogen there seem to be rays corresponding 

 to the molecule as well as to the atom. 



The author has observed rays of this type corresponding to all the elements 

 which have as yet been introduced into his tubes. These include hydrogen, 

 helium, air, carbon, neon, and mercury. Other elements are in course of 

 investigation. 



Some, but not all, of these rays have negatively charged rays connected 

 with them, resembling in this respect the secondary rays of type (2). The rays 

 from air and mercury vapour have their negative constituents, while the negative 

 rays corresponding to the hydrogen molecule, to the atoms of helium, carbon, and 

 neon, have not been detected. 



There is a considerable range in the velocities of the rays from the same 

 gas, though when the pressure is very low the greater part seem to be moving 

 with nearly the maximum velocity. 



The rays corresponding to the different atoms can be separated by deflecting 

 them by magnetic and electric forces; if after deflection they fall on a screen 

 covered with a phosphorescent substance, each kind of ray produces under the 

 simultaneous action of electric and magnetic forces a separate band on the 

 phosphorescent screen forming a kind of spectrum. The tubes I worked with 

 were not specially designed to allow the most intense magnetic fields possible 

 to be applied to the rays, and I was not sure when air was in the tube I could 

 see separate bands corresponding to nitrogen and oxygen; but when the air was 

 replaced by carbon monoxide, two separate bands, one corresponding to carbon 

 and the other to oxygen, could clearly be seen. Experiments are in progress 

 with tubes designed so that exceedingly intense magnetic fields may be applied 

 to the rays, for the purpose of using this method to analyse the gas in the tube 

 and to measure the atomic weights of the constituents. As exceedingly small 

 quantities of gas may be dealt with in this way, it appears probable that inter- 

 esting results may follow from the application of this method to the analysis 

 of the gases in vacuum tubes. 



4. The fourth type of ray is the one I have previously called 'retrograde 



