200 Dr. F. W. Aston .
Perforated Electrodes on the thick , and the second , containing ferrous iron , is but 1*5 mm. The* enormously greater absorption of the ferrous condition is shown .
A regrouping of the elements forming the exceedingly complex molecule with a tendency towards the temporary formation of ferrous salts , would go far towards an explanation .
My best thanks are due to Prof. Callendar , F.R.S. , Prof. Fowler , F.R.S. " and Prof. Conrady , for their valuable advice in the preparation of the paper. .
Experiments with Perforated Electrodes on the Nature of the Discharge in Gases at Low Pressures .
By F. W. Aston , M.A. , D.Sc .
, Clerk-Maxwell Student of the University of Cambridge .
( Communicated by C. T. R. Wilson , F.R.S. Received May 22 , 1919 .
) In a paper on the influence of the nature of the cathode on the length of the Crookes ' dark space* some difficulties in the way of a satisfactory explanation of the mechanism of that phenomenon were indicated , which showed that it would be very advantageous to obtain , if possible , direct evidence as to how much of the total current passing through the discharge tube was brought up to the surface of the cathode by the positively charged ions falling upon it .
With a view to investigating this and some kindred problems , a discharge tube was made of the form indicated in section in fig. 1 .
It consisted of a cylindrical glass shade , closed at the lower end by stout glass plate , the electrodes being flat zinc discs just filling the tube .
Zinc was used instead of aluminium as hithertofore , as , during the work with cathodes of different metals the former was found to give off very little gas and to sputter hardly at all , while the much greater ease with which it may be worked gives it a great advantage over the latter .
The upper electrode , A , was movable , being suspended by an improved form of the winch described in an earlier paper , f and which has now come into very general use in effecting accurate motion of objects within gas-tight apparatus .
* F. W. A. , ' Roy .
Soc. Proc. , ' A , vol. 87 ( 1912 ) .
t ' Roy .
Soc. Proc. , ' A , vol. 84 , p. 528 ( 1911 ) .
i Nature of the Discharge in Gases Low Pressures .
201 The lower fixed electrode , B , was perforated by a very narrowfslit , shown in section in the figure , which extended almost the whole length of its diameter .
Behind this slit , and as close to it as possible , was a Faraday cylinder , F , made of a piece of brass tube closed at both ends and opened along its length by a parallel slit a little wider than the one in the electrode , fixed as indicated and insulated from the electrode by mica .
This Faraday cylinder was protected from stray current from the gas in the space behind B by means of a guard tube , Gf , made of a half cylinder of brass closed at the ends , which completely enveloped it in a metal sheath kept at the same potential as B. Through a small hole in the guard tube an insulated lead from the Faraday cylinder was brought out through a nick in the edge of the shade , as also was a similar lead from the electrode itself .
The principal dimensions of the apparatus were as follows:\#151 ; Diameter of electrodes , 11*8 cm .
; area , 110 sq .
cm .
Length of slit , 100 cm .
Mean width of slit , determined by measuring with a high power microscope at several points , 0T82 mm. Area , 0T82 sq .
cm .
1/ 605 of the whole area of cathode .
In this apparatus the distance between the surface of the electrode and the edge of the Faraday cylinder was about 1*5 mm. This is of the order of the molecular mean free path at the pressures used , so , after some experiments had been done with it , as it was thought that so wide a gap might seriously affect the measurements of the current carried by the canalstrahlen into the Faraday cylinder , another apparatus was built , in ' which the special device indicated in fig. 2 is employed .
The lower electrode in this case was made of very stout zinc plate , the edges of the slit being chamfered off to a knife-edge as shown .
The edges of the laraday cylinder were also made of stout metal in the same way , so that it was possible to reduce the distance between the Faraday cylinder and the electrode to less than 0T mm. vol. xcvi.\#151 ; a. p 202 Dr. F. W. Aston .
Perforated Electrodes on the The following were the dimensions of the second apparatus :\#151 ; Diameter of electrodes , 11*5 cm .
; area , 103'8 sq .
cm .
Length of slit , 9*50 cm .
; area , 0105 sq .
cm .
1/ 989 total area of electrode .
As in the work on the dark space referred to ( .
cit. ) , the pressure in the tube was read by a McLeod gauge .
The current in the main discharge tube was derived from a set of accumulators controlled by a water resistance , and measured with a Hartmann and Braun milliampkremeter suitably shunted to read in fifths of a milliampere .
It was found possible to turn the apparatus to several different uses .
The following is an account of some of the experiments which were performed with it .
Experiments to Determine the Relation between the Current entering the Faraday Cylinder and that in the Main Discharge Tube .
In these experiments the lower electrode was made the cathode and connected to earth .
The Faraday cylinder was connected to a sensitive Haider DArsonval galvanometer , the other terminal of which was earthed .
The galvanometer was carefully standardised , and the scale arranged to read exactly 10 divisions per microampere .
As the resistance of the circuit was about 50 ohms and the maximum current 10 microamperes , the difference between the potential of the Faraday cylinder and that of the cathode was never appreciable .
A large number of measurements were taken with oxygen in the first apparatus , and in the second with oxygen , atmospheric nitrogen and hydrogen , the usual procedure being to set the current at a definite series of values and read off that in the Faraday cylinder corresponding to each .
Diagram I shows some of the results so obtained , the current , i , flowing into the Faraday cylinder in microamperes being plotted against the current c in the main discharge tube expressed in milliamperes .
In the set of curves , C , which were obtained with the first apparatus , the values of i have been multiplied by the fraction 605/ 989 so as to bring it into line with the other curves which were obtained with the second .
It will be seen that the curves obtained are not generally straight lines , although in the case of oxygen and nitrogen they tend to become so at low pressures , and that the curvature in the case of hydrogen is in the opposite sense to that in the other gases .
Unfortunately , on account of the very great practical difficulties involved in such a construction , the perforated electrode was not also made of the guard-ring type , * so that at low pressures , or more strictly at pressures giving * F. W. A. , ' Roy .
Soc. Proc. , ' A , vol. 79 , p. 86 ( 1907 ) .
i rzt from Faraday Cyfinder rn / Vrcno \amp ; 7TI}very# .
Nature of the Discharge in Gases at Low .
203 a long dark space , serious errors are made by the assumption that the current density on the surface of the cathode is measured by the total current passing through the tube .
Too much importance must not be therefore attached to numerical results under such conditions .
C. Oxygen ( Ftns Cult*rent rn main cf/ sch\amp ; r\amp ; e tube / n Ait)It ctmper^es .
It is particularly interesting to note that at high current densities ( in hydrogen at all current densities ) the slope of the curves does not alter very much with the pressure .
The following figures give , for currents of about 0*15 milliampere per square centimetre , the rate of change of current per unit area of the slit for a unit rate of change of current per unit area of the whole cathode .
They therefore represent the fraction of such increase of current carried up to the cathode by the stream of positively charged particles if we suppose that all falling upon the area of the slit pass into the Faraday cylinder\#151 ; fHydrogen Pressure 71 -5 45 -3 36-8 29*5 Nitrogen Fraction Pressure 0-63 22 *8 0*63 17 -6 0*64 13 -2 0*64 8*7 Oxygen ( 1st app .
) Fraction 0-40 0 43 0*45 0*50 Pressure Fraction 25 -0 0*40 17 -9 0 -46 9-2 0 *48 Oxygen ( 2nd app .
) Pressure 26 *2 21 '3 12*8 7*8 Fraction 0-43 0*48 0*53 0*55 p 2 204 Dr. F. W. Aston .
Perforated Electrodes on the From these figures , it is evident that the great reduction in the dimensions of the slit and the gap between it and the Faraday cylinder has increased the fraction of the current in the latter only by 10 per cent , or so .
From this it seems likely that the results with the latter apparatus are not very different from those which would be obtained with a theoretical apparatus in which these dimensions were infinitely small .
The results seem to indicate that about half the current in the tube is brought up to the cathode by positively charged particles having sufficient velocity to get past the slit .
A very interesting conclusion but one which must be accepted with considerable caution in the light of results quoted below on the velocity of these particles .
Experiments to Investigate the Distribution of Velocity in the Stream of Positive Particles falling upon the Cathode .
In these experiments , which were done with the second apparatus , the Faraday cylinder and its galvanometer were raised to different potentials with regard to the cathode .
It was hoped that in this way , by plotting a curve at constant pressure and current , some knowledge might be gained as to the distribution of velocity among the positive carriers , as it is clear that , if the Faraday cylinder is raised to some definite potential , Y , above the earthed cathode , only those carriers which have velocities greater than that obtained by falling through a potential , Y , will get into it at all .
On Diagram II are shown the curves obtained by plotting the results of this investigation for hydrogen and oxygen , together with the values of pressure ( P ) , current ( C ) , and voltage ( Y ) , in the main discharge tube for each curve .
It is very evident that accurate information on the point at issue is not directly obtainable in this way .
In the first place , the current reaching the cylinder is actually reversed long before the value of Y has risen as high as that between the electrodes .
, One explanation of this is to suppose that a strong current is derived from ionisation between the cathode and the edge of the cylinder ; this would also account for the very rapid change of the value of the current on both sides of the value Y = 0 .
If such an ionisation current originates entirely in the locality of the two slits , the ionisation intensity there must be far in excess of that one might be lead to infer from the consideration of the mean free path at the pressures employed .
Also , it cannot be very well due entirely to the passage of normal positively charged canalstrahlen passing between the slits , for , in one of the experiments with hydrogen , Y was increased up to the value of the potential between the Nature of the Discharge in Gases at Low Pressures .
205 electrodes ( a potential presumably sufficient to stop them completely ) , without the value of the negative current showing any signs of diminution .
D/ agkam Jl fbtenftaZ on fhractayCy/ tnchn / n Vo/ ts ( / Vydrogectbotse tAe fine , Oxygen ifj Curves for / /i/ efrogen Curves for Oxygen There seems , however , another way of looking at the matter .
This is to suppose a large number of the positively charged particles pick up , some one , some two , negative charges near the surface of the cathode .
The former , now neutral and travelling with a high velocity , might help in the general ionisation , while the latter would carry a negative current into the cylinder , unless it was at a very large negative potential .
With an increasing positive potential , on the other hand , none of the negative , but more and more of the positive , particles would be stopped , so that a reversal would take place , though it seems unlikely that it would be as great as is actually found .
From the general form of the curves , it seems probable that they represent the sum of several effects , which will be very difficult to separate from each other .
For instance , the straight line obtained for high negative potentials , which grows steeper with increase of pressure , seems to indicate that , narrow as the slit in the cathode is , the cylinder is able to create an appreciable field on the other side of it ; and so draw negative ions out of the intensely ionised region just in front of the cathode surface .
Dr. F. W. Aston .
Perforated Electrodes on the Experiments with the Perforated Electrode as Anode .
By reversing the battery and reading instruments , and so turning the lower disc into an anode , it was possible to investigate the current carried by the stream of cathode rays from the upper electrode , now the cathode , into the Faraday cylinder .
By raising or lowering the cathode , different parts of the discharge could be examined .
The measurements obtained in the negative glow were particularly interesting .
In this region it was found , using a constant current discharge , that as the distance from the cathode is increased arithmetically , the current carried by the cathode rays into the Faraday cylinder decreases geometrically .
At the edge of the dark space , the current density over the area of the slit appeared equal to that over the cathode itself , at low pressures even considerably greater .
The latter inconsistency is probably due to the fact already referred to , that the value of the current density obtained from the total , current is not reliable under those conditions , and gives readings too low by just about enough to account for it .
It seems , therefore , fairly certain that , at a point in the negative glow just beyond the dark space , practically the whole current is carried by the cathode rays , and that these rays are absorbed exponentially as they pass through the gas .
On Diagram III , the logarithm of i , the current from the cylinder is plotted against the distance between the electrodes for different pressures in oxygen .
The result is a .
series of straight lines , whose slopes appear to be a function of the pressure .
The equation connecting the values may thus be written i = ri ' or i ' e~ki , where i ' is the current carried through the slit at a point distant 1 cm .
further from the cathode than that which gives a current i. The following are values of r and k for different pressures:\#151 ; Pressure 22-0 16-0 14'2 13-4 9-1 r = 1-98 1-83 1-76 1*62 1-38 k 1-62 1-70 1-74 1-82 1-98 No very simple relation is apparent .
Two sets of readings were done in which L was fixed at 2'5 cm .
and 5*0 cm .
respectively , while the current through the discharge tube was varied .
- It was found that iincreased rapidly with c , the relation being , very roughly , c = ai^ a and b being constants depending on L. Nature of the Discharge in Gases at Low Pressures .
207 It seems probable that the determining value of L is not the distance between the slit and the cathode , but between the slit and the edge of the dark space , the latter value of L was not therefore constant in the second Diagram UI CurrenC through tube 5 rrtla mp~ i each case scale co to 2o ( origin not the same for each cur-i'c .
) set of readings , since the dark space alters in length with change of current density .
Further experiments will be done on this point when opportunity occurs .
The ones quoted above were made with the first apparatus .
Experiments on the " Anode Glow " It has long been known* that , using opposed electrodes not too far apart , when one increases the current the positive column shrinks as the negative glow stretches out from the cathode towards it , and finally disappears altogether .
With large plane parallel electrodes such as the author usually employed for work on the dark space the phenomena attending the disappearance of the positive column are very interesting .
In a gas like nitrogen the positive column shrinks very rapidly with increase of current until the last striation is left as a velvety glow on the * J. J. Thomson , ' Conduction of Electricity through Gases , ' 2nd ed. , p. 554 .
208 Dr. F. W. Aston .
Perforated Electrodes on the surface of the anode .
This appearance , which will be referred to as the anode glow , now persists hardly altered through a considerable further increase of current , and finally , when the latter does reach a certain critical value , its disappearance takes place with surprising suddenness .
In 1906 when the conditions best suited for the measurement of the dark space were being examined it was noticed that the value of the critical current could be determined with considerable accuracy .
The fact that in the second discharge tube , described above , the distance between the electrodes could be altered with great accuracy from 0 cm .
to 20 cm .
offered a very favourable opportunity for determining whether the disappearance of the anode glow was as definite a function of the distance from anode to cathode as it was of the current passing between them .
On trying the experiment remarkable results were obtained .
In the discharge in pure atmospheric nitrogen at a pressure of about 0*15 mm. , with a dark space of 1 cm .
, it was found that a movement of one of the electrodes of less than a millimetre was quite sufficient to determine the appearance or disappearance of the anode glow .
And in general whatever the pressure and current , so long as they were constant , the distance between the electrodes at which the anode glow just disappeared could be determined to within 1 per cent. A series of measurements were therefore carried out with nitrogen at various pressures , the procedure being as follows:\#151 ; The current was set at a definite value .
The upper electrode was then moved until the critical point was reached , it was then lowered a little to ensure complete disappearance .
The current was now set exactly and the voltage read off .
The upper electrode was then raised very carefully until the glow reappeared , the distance between the electrodes was then read off on a millimetre scale .
When these results were plotted it was soon recognised that the important factors to be measured were the voltage and the distance between the anode and the edge of the dark space .
A second series of experiments were done paying special attention to these , the results of which are plotted on Diagram IY .
It will be seen that at constant pressure L , the distance between the anode and the edge of the dark space is a function of Y , the potential difference between the electrodes ( which has been already shown to be identical under these conditions to the fall of potential across the dark space)* becoming almost linear at low pressures .
And at constant Y it is very nearly inversely proportional to the pressure .
* ' Roy .
Soc. Proc. , ' A , vol. 84 , p. 526 ( 1911 ; .
i Nature of the Discharge in Gases at Low Pressures .
209 The following are the values of L and P at constant potential of 500 volts : p 21-0 18-4 15-7 11-9 L 11-4 13-0 15-0 19-0 L x P ... .
240 240 236 229 If the lower perforated electrode'is made the anode , the appearance of the glow is invariably attended by a sudden jump in the value of the negative current delivered into the Faraday cylinder .
This is , indeed , a very delicate means of determining the critical point .
The potential between the electrodes also always increases at the same time by a few volts .
The glow must be as intimately connected with the surface of the anode as it is with the distance from the cathode , for not only does it appear brighter on some parts of the anode than others , but also , once it has appeared , the distance between the electrodes can be increased enormously without altering any factor in the discharge appreciably .
The critical point would seem to be that at which the energy in the stream of cathode rays falls off to a definite minimum intensity .
As this energy is almost certainly a function of both the potential and the current , the influence of these separately will have to be determined , e.g. , by the use of cathodes of various metals , before anything definite can be stated .
It is interesting to note that the spectrum of the anode glow is widely different from that of the negative glow in nitrogen , the former consisting entirely of bands , the latter largely of lines .
Further experiments are projected on this phenomenon .
210 On the Nature of Discharge in Gases at Low Pressures . !
In conclusion , I should like to express my thanks to Prof. Sir J. J. Thomson for the interest he has shown in these experiments .
Summary. .
Experiments are described on the discharge between electrodes of a large flat form perforated with a long narrow slit , the charge passing through the slit being collected and measured in a Faraday cylinder .
Direct measurements , made with the Faraday cylinder behind the cathode and at the same potential , seem to indicate that about half the total current in the discharge is brought up to the cathode by positive ions .
Attempts to discover the distribution of velocities in this stream show that this is not directly determinable , owing to the very high ionisation in the region of the slit and other reasons which are discussed .
Using a perforated anode , it is found that , as the distance from the cathode is increased arithmetically , the current carried by the cathode rays into the Faraday cylinder decreases geometrically when the current is constant .
The effect of { a ) the distance apart of the electrodes and ( 6 ) the total current flowing , on the disappearance of the last trace of the positive column or anode glow , is shown to be remarkably definite , but no adequate explanation of this phenomenon seems available at present .

