468 



NATURE 



[June i, 191 i 



(i) Ordinary molecular oxygen, O,. 



(2) Neutral atoms of oxygen, O. 



(3) Atoms of oxygen with i positive charge, 0+. 



(4) Atoms of oxygen with 2 positive charges, O ++. 



(5) Atoms of oxygen with 1 negative charge, 0_. 



(6) Molecules of oxygen with i positive charge, 0,+. 



(7) Ozone with a positive charge, O, +. 



(8) O, with a positive charge, O,. 



And, in addition, there are free negative corpuscles. 

 Thus in the elementary gas there are at least nine (the 

 list has no claim to be exhaustive) separate substances 

 present when the discharge passes through it. Each of 

 these substances has almost certainly different properties, 

 possibly 3 characteristic spectrum. If we took any other 

 gas wo should find that the same thing would be true : 

 thus in hydrogen we have H, H,, H +, H_, Ho+j even 

 if we do not asvribe to hydrogen the lines giving mje = i-^ 

 or 3. In nitrogen we have N, N,, N+, N++, fi^+t 

 carbon occurs as C+. C++, C_, chlorine as CI, Clj C1+, 

 C1j.+, and Cl_, mercury as Hg, Hg+, Hg++, and prob- 

 ably as Hg +++, as there is a very persistent line for 

 which tn/e is about 66. 



Thus, whenever the electric current passes through a 

 gas, and probably whenever a gas is ionised, the gas be- 

 comes a mixture of many different substances. We can 

 thus readily understand why in the spectra of many 

 elements many of the lines may be grouped together so 

 as to form different series — the principal series, the first 

 coordinate series, and so on — and the spectrum of the dis- 

 charge tube regarded as the superposition of a number 

 of different spectra the relative intensities of which may 

 be subject to vory rtront variations. This, indeed, is just 



Fig. 4. 



what would happen if some or all of the substances which 

 are present when the gas is in the ionised state gave rise 

 to different spectra. 



Another feature which I think is of great interest from 

 the point of view of the theory of chemical combination 

 IS the occurrence of particles with negative charges. Let 

 us consider for a moment how these are formed. They 

 are formed after the particles have passed through the 

 kathode; the path between the kathode and the photo- 

 graphic plate contains abundance of corpuscles produced 

 by the lonisation of the gas ; a neutral particle, after pass- 

 mg through the kathode, picks up a negative corpuscle, 

 and so becomes negatively charged. For this to occur, the 

 attraction between the corpuscle and the neutral particle 

 must be exceedingly strong, for it is not a question of a 

 particle at rest attracting to itself a negatively electrified 

 corpuscle sauntering about in its neighbourhood. In our 

 case the neutral particle is rushing past the corpuscle with 

 a velocity of the order of io« cm. per sec. In order that 

 the particle may in these circumstances be able to drag 

 the corpuscle along with it, the attraction between the two 

 must be so great that to move a corpuscle against this 

 attraction from the surface of the particle away to an 

 mhnite distance must require an amount of work of the 

 same order as that required to communicate to the cor- 

 puscle a velocity of 10' cm. per sec. ; this is equal to the 

 "^o^^ JJ^^ired to move the atomic charge against a poten- 

 tial difference of about 3 volts, and is therefore comparable 

 with the work required to dissociate some of the most 

 stable chemical compounds. 



The fact, then, that some particles get negatively 

 charged shows that in the neutral state these particles 

 have an exceedingly strong affinity for a negatively 

 NO. 2170, VOL. 86] 



electrified particle, while the absence of a ]> 

 particle from the negative side shows that its a;. 

 much less, but does not imply that it vanishes altuget 

 From what we have said, it should folk>w that the n 

 slowly the neutral particles are moving relatively to 

 corpuscles, the more easily will the negatively electrif 

 systems be formed. This is confirmed in a very strikii 

 way by our experiments, for when the discharg',- is passit 

 very easily through the tube, and the velocity of 

 neutral particles is relatively small, the number of n^i 

 tively electrified particles is very much increased ; inde 

 in some cases the brightness of the part of the ph 

 graph corresponding to the negative particles is as gre 

 as that corresponding to the positive, whereas when th«' 

 discharge is passing with great difficulty, and the velocity 

 of the neutral particles is very high, the negative part is 

 very faint compared with the positive. 



The particles which have been observed on the ne^jativo 

 side are the hydrogen atom, the carbon atom, th' 

 atom, and the chlorine atom. The presence of ox; 

 chlorine atoms might, perhaps, have been expected, 

 are universally regarded as strong electro-negative < 

 i.e. as elements which have a strong affinity for ; 

 electricity. The presence of the hydrogen atom is nior- 

 remarkable, for hydrogen is generally considered to be a 

 strongly electro-positive element, yet on these photographs J| 

 we find it more persistently on the negative side than anvv 

 other particle ; often when no other line on the r-"-^'--" ^ 

 side is strong enough to be detected, the line coi 

 ing to the hydrogen atom is distinctly visible. Th ; 



the more remarkable, because the hydrogen atom, bein| 

 the lightest of all the particles, is moving with the great 



Fig. 5. 



velocity relatively to the corpuscles, and therefore would, 

 other circumstances being the same, be the least likely ta 

 capture them. The heavier the particle, the slower is its 

 velocity, and the greater chance it has of capturing the 

 corpuscles ; the fact that heavy complicated particles ar9| 

 conspicuous by their absence on the negative side showi 

 that the attraction of these for the corpuscles must 

 exceedingly small compared with that exerted by a neutr 

 atom of hydrogen. It will be seen that the atom ofl 

 carbon, also regarded as an electro-positive element, is also^ 

 conspicuous on the negative side. 



On looking at the list of the particles which occur 

 the negative side, we are struck by the fact that they ar 

 all atoms : there is not a molecule among them. Thus 

 although the curve corresponding to the negatively elec-^ 

 trifled hydrogen atom occurs on every plate, there is nc 

 a single plate which shows a trace of a curve correspond* 

 ing to a negatively electrified hydrogen molecule, althougl 

 that corresponding to the positively electrified molecule 

 always present, and on some of the plates is stronger ths 

 that due to the positive hydrogen atom. Again, on son 

 plates the positive oxygen molecule shows stronger tht 

 the oxygen atom, but on the negative side only the at 

 is visible. 



Thus neutral atoms," but not neutral molecules, 

 exert on the negative corpuscles those enormous att 

 tions which, under the conditions of these experiments, 

 required to bind the corpuscles to these rapidly movit 

 particles. We may compare this result with the propertied 

 ascribed by chemists to bodies when in the nascent con- 

 dition, i.e. when they have only recently been liberated 

 from chemical combination, and when they are likely to 

 be partly in the atomic state, for atoms, as we have seen. 



