23° 



NA TUBE 



[July 3, 1884 



Very often direct observation has shown the existence of these 

 •dusts in drops of rain, and this is what has happened in all 

 parts of the world since the crepuscular colorations of 18S3-84. 

 The dusts collected have a composition which usually indicates 

 a volcanic origin. It has been shown that other volcanic erup- 

 tions have been followed by red glows in the sky ; it appears to 

 me that it may also be shown that they have been followed by 

 abundant rains. The eruptions which have been referred to 

 are those of the Skaptar Joekull, in Iceland, in the beginning of 

 May 1783 ; of a new volcano, since disappeared, in the Sicilian 

 Sea early in July 1831 ; Cotopaxi, in America, in 1856 ; 

 Vesuvius in 1S62. These eruptions were followed by colora- 

 tions ; I add that they were followed by rains which exceeded 

 the mean. The following, in millimetres, are the monthly heights 

 of rain collected on the terrace of the Paris Observatory; the 

 second line is the monthly mean of from twenty to thirty 

 years : — 



May rune July Aug Sept. 



'7S3 62 86 43 75 51 



Means 47 49 86 47 42 



Da. Nov. Dec. Inn. 1S32 



lS 3' 52 76 J6 35 



Mean -< 41 47 34 34 



\pril May June luly Aug. Sept. 



l8 5° 51 117 49 54 54 60 



Me:lns 37 53 54 55 45 4§ 



EXPERIMENTS ON THE PASSAGE OF ELEC- 

 TRIC/TV THROUGH GASES—SK'ETCH OF 

 A THEORY' 



'TTIh. passage oi electricity through gases has of late years 

 become a very favourite subject for experimental investiga- 

 tion. A large number of frets have thus been accumulated, 

 and it becomes of importance to see whether these facts throw 

 any light on the theoretical notion, which we have based on 

 other branches of electrical inquiry. 



If we have two bodies at a different electrical potential sepa- 

 rated by a layer of air, we might imagine the air in contact witli 

 the bodies to become - lectrified, then move on, impelled by the 

 electric forces, and re-establish equilibrium by giving up their 

 charges. The passage of electricity through gases would then lie 

 similar to the diffusion of heat. But, however natural such a 

 view would be, it is impossible to maintain it in the face of ex- 

 perimental facts. The experiments which I shall bring before 

 you to-day seem to me to »upp >rt, on the contrary, the idea that 

 the passage of electricity through a gas resembles the phenome- 

 non studied by Helmholtz under the name of electrolytic con- 

 vection. 



I shall avoid as much as possible all suppositions and hypo- 

 theses which cannot be put to the test of experiment ; but it 

 seems necessary to start with some assumption in order to avoid 

 too great a vagueness in the subsequent explanations. The 

 assumption which I shall make is this : In a gas the passage of 

 electricity from one molecule to another is always accompanied 

 by an interchange of the atoms composing the molecule. I 

 shall also try to prove that many facts are easily explained by 

 1 lie assumption that the molecules are broken up at the negative 

 pole. 



If, in a vacuum-tube of the ordinary form, the discharge is 

 passed at a pressure of about one millimetre, a luminosity is seen 

 round the negative pole which is called the negative glow. A 

 luminous tongue projects from the end of the positive pole, 

 which I shall call the positive part of the discharge, without 

 meaning to imply that it is charged with positive electricity. 

 The positive part of the discharge and the negative glow are 

 separated by a non-luminous space, which I shall call "the 

 lark interval." The glow itself is divided into three layers, the 

 thickness of which increases with decreasing density. Closely 

 surrounding the electrode itself, we have in the first place a 

 luminous layer, which on new electrodes is of a golden colour. 

 The spectroscope shows the presence of sodium and hydrogen ; 

 the sodium is due to foreign matter deposited on the electrode, 

 and the hydrogen is expelled by the action of the heat out of the 



1 Abstract of ihe ItiUnan I . . lure. Read before the Royal Society, 

 June 19. 1884, by Arthui Si huster, Ph.D., F.R S. 



electrode by which it had been absorbed. When the electrodes 

 have been in use for some time, the golden colour disappears, 

 and the spectrum belonging to the gas used is seen. The second 

 layer is known by the name of the dark space. The third layer 

 is the glow proper. 



The theory which I shall endeavour to establish is this : That 

 within the first layer the gaseous molecules are decomposed, 

 that their negative parts are projected with great velocity 

 through the dark space, that this velocity is gradually reduced 

 by impacts within the glow, and that in the positive part of 

 discharge the discharge takes place by diffusion except when 

 stratifications appear. 



According to the kinetic theory of gases, the molecule of 

 mercury vapour consists of a single atom, which is incapable of 

 vibration. Mercury lias a very brilliant spectrum, which proves 

 that the theory is incomplete in some important point. It is 

 well known, on the other hand, that the theoretical conclusion 

 receives support from the fact that the vapour-density of mercury 

 vapour is anomalous. If, as is generally supposed, the molecule 

 of the majority of gases contains two atoms, that of mercury can 

 only contain one. If an essential part of the glow discharge 

 is due to the breaking up of the molecules, we might expect 

 mercury vapour to present other and much simpler phenomena 

 than the gases with which we are generally accustomed to work. 

 This, indeed, is the case ; for I find thai, if the mercury vapour is 

 sufficiently free from air, the discharge through it shows no nega- 

 . no dark sfua s, and no stratifications. At the ordinary 

 temperature the spark does not pass through mercury vapour; 

 but il a tube free of air, but containing mercury vapour, is 

 heated, the discharge passes always in a continuous stream of 

 light. It is not always quite symmetrical with respect to the two 

 poles ; and a very curious tendency of the spark is noticed, to 

 pass at the negative pole rather from the glass out of which the 

 electrode protrudes than from the metallic electrode itself. A 

 brilliant sodium spectrum then appears at the point from which 

 the spark sets out. Whenever small traces of air remain, strati- 

 fications are very apt to appear, as a mixture of air and mercury 

 gives fine stratifications, but I have never noticed them after 

 sufficient removal of the air. 



I now pass to the description of an experiment which seems 

 to me to be only capable of explanation by the views brought 

 forward in this paper, and I should like therefore to consider 

 them as crucial experiments, which have to be explained by any 

 true theory of the discharge. As negative electrode, I use an 

 aluminium cylinder of 5'5 cm. internal diameter and 8 cm. long. 

 A long aluminium wire running parallel to the axis of the cylin- 

 der at a distance of about an inch formed the positive electrode. 

 On exhaustion, the discharge at first passes as a spark in the 

 ordinary way, but as the pressure decreases the glow gradually 

 surrounds the whole cylinder, with the exception of a dark strip 

 about 2 or 3 cm, in width, directly oppositt the positive wire. The 

 positive electrode seems, therefore, to repel the negative glow. 



The following seems to me a plausible explanation of the 

 phenomenon which I have just described. The rapid fall of 

 potential which is observed on crossing the negative electrode 

 suggests at once, independently of any theory that we have to 

 deal with, the action of a condenser, for we know that no stati- 

 cal charge can produce a finite difference of potential at the 

 electrode, while a double layer will produce a discontinuity. 

 Although it may not be proved that an absolute discontinuity of 

 potential exists at the kathode, it is yet certain that a very rapid 

 tall occurs at that place. This is all that is necessary for the 

 argument. 



We recognise such a double layer in the case of electrolytes, 

 but there is an essential difference in the thickness of the layer 

 within which we must imagine that condenser action to take 

 place. In the liquids that thickness must be very small, as is 

 shown by the intensity of the observed polarisation currents. 

 The positively electrified matter in every case is kept against the 

 negative surface by a joint action of electrical and chemical 

 forces, for it has been shown by Helmholtz that only thus can 

 we explain a difference of potential between two bodies. It is 

 the chemical forces which keep the electricities asunder. The 

 gaseous molecules or atoms, however subject to their mutual 

 encounters, and always having certain velocities, will tend to 

 leave the surface. They are kept near the surface, however, by 

 the electrical forces. 



Suppose, now, that a positive electrode is placed near such a 

 condenser. The resistance of the gas is so much greater than 

 that of the metal electrode that we shall assume the whole elec- 



