858 Mr. G. P. Thomson on the Application of 



a dark space of from about '5 to 5 cm., a beam of coloured 

 light is seen to diverge from the surface of the anode. 

 This beam is caused by the passage of anode rays, which are 

 then analysed by being allowed to pass through a narrow 

 tube of '2 mm. diameter and 4" 5 cm. length, thus separating 

 out a fine pencil which is subjected to electrostatic and 

 magnetic deflexions. 



Although the chief interest in these experiments has been 

 the nature of the rays rather than the mechanism by which 

 they are produced, it may not be out of place to give some 

 details of the methods which have been found necessary in 

 practice to produce a satisfactory beam of rays. 



The first requisite is that the powder should be as good a 

 conductor as possible. To ensure this it should be finely 

 ground and rammed into the tube as firmly as possible : 

 indeed, considerable force should be required to dislodge it 

 when the anode is eventually taken to pieces. Gehrcke and 

 Reichenheim appear generally to have heated their mixtures. 

 This is of advantage in certain cases, especially with easily 

 melted salts, where the melting will improve the mixing. 

 It is not, however, indispensable, and in certain cases is 

 inadmissible owing to the decomposition produced (e. g. of 

 BeF 2 ), while in many others it has little effect. The 

 presence of a halogen is apparently essential to the pro- 

 duction of anode rays by this method, and of the halogens 

 iodine appears to be best, and then bromine. The function 

 of the halogen is undoubtedly to produce an abnormally 

 large anode fall of potential by absorbing electrons and so to 

 cause a vigorous bombardment of the anodes ; it is probably 

 immaterial from what source it is liberated into the dis- 

 charge-lube. 



The metals examined differed considerably in the ease 

 with which results could be obtained from them, lithium 

 being the easiest and strontium the hardest. Thus the 

 relative strength of the parabolas due to two metals in 

 the rays from a mixture of salts is not given by the relative 

 number of atoms of the two present in the mixture. The 

 strength of the parabolas decreased with increasing atomic 

 weight in each of the two chemical groups examined. This 

 is partly accounted for by the fact that the photographic 

 plate is less sensitive to the slower-moving heavier-particles, 

 but there seems also to be a decrease in the brightness of 

 the light produced in the beam, and it is probable that there 

 are actually fewer of the rays of high atomic weight. 

 The elements of the alkaline earths tend to give weaker 

 ravs than the corresponding alkalis, and it is sometimes 



