Rays of Positive Electricity. 235 



indication o£ secondary radiation' of this atomic weight when 

 no helium was present, that I was inclined to doubt whether 

 the hydrogen molecule ever did occur as secondary radiation; 

 quite recently, however, a tube which had got contaminated 

 with grease from one of the taps showed the molecule of 

 hydrogen more strongly than the atom in the secondary 

 radiation ; a photograph taken with this tube is shown in 

 fig. 14. There must therefore, I think, be some types of 

 compounds of hydrogen from which the hydrogen molecule 

 is liberated to form the greater part of the secondary radia- 

 tion due to hydrogen, though in hydrogen itself and in 

 numerous other compounds of hydrogen it is the hydrogen 

 atom which constitutes the larger part of the secondary 

 radiation. It is remarkable that, even in a case like that 

 shown in fig. 14, where the positively-charged molecule 

 largely predominates in the secondary radiation over the 

 positively-charged ion, the negatively-charged hydrogen 

 molecule cannot be detected, while the effects of the nega- 

 tively-charged hydrogen atom are quite distinct. 



In many tubes I have observed a secondary radiation for 

 which the electric atomic weight is about 1*4 ; the curves on 

 the plate due to it are faint, but they occur with a great 

 variety of gases, and with negative as well as positive 

 charges. 



Another secondary radiation, stronger than that just men- 

 tioned, but not often very conspicuous, is one with an electric 

 atomic weight of 3 ; it is specially bright when hydrogen is 

 in the tube, while carbon compounds do not exhibit it to any 

 very marked extent. The electric atomic weight would fit in 

 either with a complex ion H 3 with one charge, or an atom 

 of carbon with four charges. Since it seems not to be 

 specially increased by the introduction of carbon compounds, 

 I think the former hypothesis is the more probable. It occurs 

 with a negative as well as with a positive charge. The 

 radiation mentioned in the preceding paragraph ma}' be due 

 to this ion with a double charge. 



Another type of secondary radiation, much stronger than 

 those just mentioned and especially bright in carbon com- 

 pounds, is one with an electric atomic weight of 6 ; both 

 this and the preceding type are shown in fig. 15. I have 

 not observed this unless carbon or some of its compounds 

 are present, and so I regard this as due to a carbon atom 

 with a double charge. It is only found with a positive 

 charge. 



In compounds not containing carbon there is secondary 

 radiation with an electric atomic weight verv near the last, 



