333 An Investigation of the Ionising Power of the Positive Ions from a Glowing Tantalum Filament in Helium .
By Frank Horton , Sc. D. , Professor of Physics in the University of London , and Ann Catherine Davies , M.Sc .
, Royal Holloway College , EnglefieM , Green .
( Communicated by Mr. C. T. R. Wilson , F.R.S. Received August 16 , 1918 .
) The ionising power of the positive ions emitted by a glowing solid has been investigated by Pawlow* and by Bahr and Franek.f The method used by these observers was that devised by Lenard for determining the ionising potential for electrons , and subsequently used by several experimenters for the same purpose .
The positive ions from a glowing filament were accelerated towards a piece of platinum gauze by an electric field which could be varied .
Those ions which penetrated through the gauze encountered a retarding electric field , greater than the accelerating field by a constant amount , which prevented any of them from reaching a collecting electrode placed a few centimetres beyond the gauze .
If , however , any of the positive ions attained sufficient velocity to ionise the gas molecules on the further side of the gauze , the electrons produced would be attracted towards the collecting electrode , and their presence would be indicated by a deflection of the electrometer connected to it .
The experiment consisted in gradually increasing the accelerating field until the presence of electrons beyond the gauze was indicated by the electrometer .
Pawlow came to the conclusion that the ionising potential for positive ions in hydrogen was about 10 volts , but the results with the other gases tested were inconclusive and were taken to indicate that the ionising potentials were lower the hotter the source of the positive stream .
Bahr and Frank experimented with air and with hydrogen , and could not obtain any definite evidence of the existence of a minimum ionising voltage tor positive ions in either of these gases .
They found that the point at which electrons could first be detected in the ionisation chamber varied with the density of the positive stream , and they concluded from their results that the gases were being ionised by the positive ions , and that , if there is a definite minimum ionising potential , it is lower than the ionising potential in the case of electrons .
The authors have been tor some time conducting experiments involving the ionisation of helium gas by collisions , and in the course of this work were led * W. I. Pawlow , ' Roy .
Soc. Proc. , ' A , vol. 90 , p. 398 ( 1914 ) .
t F " v. P.ahi and J. Frank , * Deutseli .
Phys. Ges .
Verh .
, ' vol. 16 , p. 57 ( 1914 ) .
Prof. F. Horton and Miss A. C. Davies .
to make a direct investigation of the ionising power of the positive ions emitted by a glowing tantalum filament .
An account of these experiments is given in the present paper ; they lead to the conclusion that the minimum ionising potential for positive ions is very much greater than that for electrons .
Description of the Apparatus and the Method of Experimenting .
The method used is , in principle , that which has already been described , but a few modifications in detail have been introduced .
A diagrammatic representation of the apparatus is shown in fig. 1 .
The source of positive ions was a short piece of tantalum filament , G , which could be heated by an electric current supplied I through the stout platinum leads , E and F. The ionisation \ chamber , B , consisted of an aluminium tube , lined inside TO *---^e0cm , r with thin platinum foil .
The lower end of the tube was covered by a piece of fine platinum gauze ( 100 wires to the inch ) , stretched tightly across the end and held in position by an aluminium ring .
The surface of the gauze was plane , and parallel to the platinum plate A , which served as the collecting electrode .
This was suspended inside the cylinder by a fine platinum wire , the upper end of which was tied to a silk thread wrapped round the plug of a stopcock adapted to this purpose .
By turning the stopcock , the distance between the collecting plate and the grid could be varied from zero to 6 or 7 cm .
The top of the ionisation chamber was nearly closed by a circular platinum disc with a hole in its centre to allow of the free passage of the wire suspending the plate A. Above this cover the wire passed through a small platinum ring , making electrical contact with it , and by this means the collecting electrode was connected to the quadrant electrometer used to measure the ionisation produced .
The inside of the ionisation chamber was thus entirely of platinum , and the outer aluminium cylinder served .merely as a support for the platinum foil which formed the walls , and for the thin platinum gauze which covered its lower end .
The metal parts of the apparatus were contained inside a glass tube of about 3'5 cm .
internal diameter , which the aluminium cylinder just fitted .
The glass tube was about 21 cm .
in length , and was closed at the top by a glass bulb , which carried the stopcock by means of which the position of the collecting electrode could be adjusted .
The bulb was fixed to the top of the Fig. 1 .
An Investigation of the Ionising Power of the Positive Ions .
335 tube by sealing-wax , applied so as to have as small a quantity of this as possible inside the apparatus .
The sealing-wax joint was 8 cm .
above the top of the ionisation chamber , and was completely shielded from radiation from the filament ; it did not become heated during the experiments .
All the platinum used in the apparatus was boiled for many hours in strong nitric acid before being fitted into position , and the whole apparatus was finally washed out wit*h chromic acid , followed by distilled water , in order to remove traces of grease from handling .
The treatment of the platinum with nitric acid is that usually adopted as tending to prevent the formation on its surface of electrical polarisation layers , which are a source of trouble in experiments of this nature .
The helium gas used in the experiments was purified by allowing it to stand for several hours over charcoal cooled in liquid air .
The purified gas was stored in a bulb of about half a litre capacity , which communicated through a stopcock with a long fine capillary tube , which allowed the gas to enter the exhausted apparatus at a very slow rate .
The gas entered through a U-tuhe .
one end of which was fused on to the glass tube containing the ionisation chamber , at the level of the filament , and the other to the capillary tube .
The IJ-bube contained carbon and was cooled in liquid air during the experiments .
The limb of the U-tube connected to the ionisation chamber was about 1'3 cm .
wide , in order to give free communication between the ionisation chamber and the purifying apparatus .
This was important as some observations were made at a constant pressure , and in these cases a stream of gas was not used .
The glass tube containing the ionisation chamber was also connected to a drying tube , a McLeod gauge , a mercury vapour pump , and a wide tube with a mercury column , the height of which could be adjusted so as to alter the volume of the evacuated space and so vary the pressure of the residual gas .
The tube containing the ionisation chamber could be cut off from the rest of the apparatus by a stopcock , on the pump side of which was a narrow glass U-tube immersed in solid carbonic acid , and on the other side a tube , 10 cm .
long , filled with gold leaf .
These served to prevent mercury vapour from passing over into the ionisation chamber .
After the apparatus had been connected to the pump and evacuated , a large asbestos box was built up round the lower part of the glass tube containing the ionisation chamber .
The box was open at the bottom and a Bunsen flame was placed underneath .
In this way the whole of the metals contained in the experimental tube was heated to about 250 ' C. The charcoal tube ( for cooling with liquid air ) was strongly heated and the vapour pump was worked continuously .
When a high vacuum had been established , the Prof. F. Horton and Miss A. C. Davies .
tantalum filament was raised to a white heat by an electric current,.and the heating of the filament and the apparatus was continued for many hours until the McLeod gauge indicated that the evolution of gas had practically ceased .
The temperature of the filament during the experiments was read by means of an optical pyrometer , * but the process of driving the gas out of the filament by long-continued heating at a high temperature had caused a slight blackening of the glass tube through which the filament was viewed , and its temperature was probably much higher than that indicated .
The optical pyrometer readings recorded throughout the paper are , therefore , probably considerably lower than the actual temperatures of the filament , but the pyrometer was a useful means of obtaining an estimate of the difference of temperature in two experiments , and also of maintaining the temperature of the filament constant throughout a series of observations at different gas pressures .
In applying Lenard 's method to investigate the ionising power of electrons or positive ions from a glowing solid , it has been usual to arrange the circuits so that the retarding potential inside the ionisation chamber is always a few volts greater than the accelerating potential below the gauze .
The retarding field thus increases as the accelerating field is increased , and this method has the disadvantage that , if there is any other cause producing ionisation inside the ionisation chamber , the number of ions so produced which reach the collecting electrode may increase as the retarding field is increased , and this spurious effect may be mistaken for that which the experiment is designed to detect .
In order to get a copious supply of positive ions from a hot filament which has been in use for some time , it is necessary to maintain it at a very high temperature , and a considerable amount of light is consequently produced inside the apparatus .
This light illuminates the gauze and shines through the interstices upon the walls of the ionisation chamber , which , like the gauze , are at a negative potential , and under the influence of the light emit photo-electrons , which pass to the collecting electrode .
In the present experiments this photo-electric effect was always very considerable , especially when the filament was at the highest temperatures used .
It does not seem to have been observed by Pawlow or by Bahr and Frank , and it was no doubt smaller in their experiments , because the negatively charged wall of the ionisation chamber was made entirely of gauze , so that a smaller metal surface was exposed to the radiation .
In the experiments now described the photo-electric current between the walls of the ionisation chamber and the platinum disc suspended inside increased considerably as the difference of * The optical pyrometer was purchased by a grant from the Government Grant Committee of the Royal Society .
An Investigation of the Ionising Power of the Positive .
337 potential between them was raised , especially at certain pressures .
For this reason sets of observations were always made with a constant field in this space , the potential difference between the collecting electrode and the gauze being in each experiment somewhat greater than the largest value of the accelerating potential which was used in that series of observations .
In order to test the constancy of the photo-electric effect during a series of observations , these were always taken both with increasing and with decreasing accelerating potentials .
The accelerating potential was first gradually increased to its maximum value , and then gradually decreased .
The curves obtained when the results were plotted were usually similar in form , but not quite coincident .
This may have been due to an alteration in the photo-electric effect or to a change in the thermionic emission .
In some cases the accelerating voltage was increased and decreased alternately many times , and the later values obtained were usually in good agreement Preliminary Experiments .
In the first experiments the helium used was not quite pure .
It had been carefully purified by standing over carbon cooled in liquid air before being stored , but liquid air was not available for cooling the carbon tube connected to the apparatus during the experiments , and this tube was surrounded by solid carbonic acid .
This arrangement was not nearly so efficient as when liquid air was used , and unless a stream of helium was kept flowing through the ionisation chamber , the helium probably became contaminated with gases evolved from the heated filament .
However , the amount of impurity could never have been large , and the results obtained were in general agreement with those obtained later when pure helium was used .
The results of a set of observations taken under these conditions at a pressure of 0-069 mm. are graphically represented in fig. 2 .
In this set , the platinum collecting electrode was 2 cm .
above the gauze , and between these a constant potential difference of 8o'4 volts was maintained .
The accelerating voltages were measured between the positive terminal of the filament and the gauze , and there was a difference of potential of about 0'7 volt between the terminals of the filament .
With an accelerating potential of only a volt or two , the filament gave a large negative emission .
About 4 volts were necessary to prevent this negative emission , and as the accelerating potential was further increased , the platinum disc collected electrons at a rate which was independent of the velocity of the positive ions until an accelerating potential of about 20 volts was reached , after which the rate of collection increased uniformly with the accelerating potential as shown in the figure .
When a maximum of 66 volts had been attained , the accelerating field was Prof. F. Horton and Miss A. C. Davies .
gradually decreased .
The currents measured by the electrometer with decreasing voltages were all very slightly larger than those obtained when 10 20 30 40 .
50 60 70 POTENTIAL DIFFERENCE Fig. 2 .
the field was being increased .
The part of the curve indicating a constant current in the ionisation chamber no doubt represents the photo-electric emission from the walls of the chamber under the influence of the light from the glowing filament .
\#166 ; The break in the line at 20 volts indicates the production of electrons in the ionisation chamber by some other means , which begins when the accelerating potential difference has reached that value , and increases steadily with that potential difference .
The curve is typical of several which were obtained under conditions similar to those described , and is of the form to be expected if the positive ions from the glowing tantalum , after being accelerated by a potential difference of 20 volts , have sufficient energy to ionise the helium atoms on collision with them .
The usually accepted value of the minimum ionising potential for electrons in helium is 20 volts , and if the ionising potential for positive ions in helium were the same , the result would be similar to that obtained by Pawlow for hydrogen .
In his experiments with helium , Pawlow found that the point at which the curve began to rise depended on the- temperature of the filament , the rise beginning at a lower accelerating potential the higher the temperature .
A similar result was obtained by Bahr and Frank with other gases .
The temperature of the filament during the observations recorded in fig. 2 was read by the optical pyrometer as 1500 ' C. In order to see whether the point at which the curve begins to rise depends on this temperature , the experiment was repeated with the filament at about 1400 ' C. At this lower temperature the filament was less luminous , and the positive emission much smaller .
The negative currents measured by the electrometer were all An Investigation oj the Ionising Power of the Positive Ions .
33 !
\gt ; smaller than before .
The break in the curve still occurred at about 20 volts , but the slope with increasing accelerating potential was less steep than when the higher temperature was used .
The results of these experiments agree with the view that there is a. definite minimum ionising potential for positive ions in helium of about 20 volts , but other possible explanations of the results can be suggested .
In the first place , the agreement with the value which is usually taken to be that of the minimum ionising potential of electrons in helium , suggests that the result may be due in some way to electrons .
The gradual increase of the accelerating potential below the gauze cannot directly influence the electrons in the ionisation chamber , because of the very perfect shielding of the fine gauze .
The increased supply of electrons must either come through the gauze from below , which is unlikely seeing that the electric field opposes them , or they must be produced by the positive ions which get through the gauze , or by a radiation which originates below the gauze and passes into the ionisation chamber .
The surface of the gauze facing the filament is strongly illuminated by the light of the filament , and as the gauze is negatively charged , photo-electrons are no doubt emitted and stream towards the filament.* When the accelerating potential reaches 20 volts the electrons lose energy by collisions with the helium atoms which are usually assumed to be ionised by such impacts .
If , however , the collisions result in the production of radiation ( whether accompanied by ionisation or not ) , this radiation might penetrate into the ionisation chamber and there cause the production of electrons , either by directly ionising the helium or by causing an increased photo-electric emission from the chamber walls .
If the electron collisions produce ionisation only , unaccompanied by radiation , at 20 volts , then , when the accelerating potential reaches this value , there will be an increased supply of positive ions passing into the ionisation chamber , but there will be no extra supply of electrons inside the chamber unless the positive ions give rise to them .
It does not , however , necessarily follow that the positive ions ionise the gas .
They collide with the helium atoms , they are repelled from the platinum collecting electrode , and they ultimately collide with the gauze or the walls of the ionisation chamber .
It is well known that when a negatively charged metal plate is bombarded by positive ions of sufficiently high velocity it emits electrons , but the velocities at which this effect has been observed hitherto have been much greater than those csed in the present experiments .
If , however , electrons are emitted from * The recent experiments of Richardson and Bazzoni ( ' Phil. Mag. , r VI , vol. 34 , p. 285 ( 1917 ) ) make it appear improbable that a radiation is produced by the electrons bombarding the filament .
Prof. F. Horton and Miss A. C. Davies .
platinum when bombarded by positive ions with a velocity of 20 volts , and if the number of electrons emitted increases with the velocity of the impinging positive ions , the form of the curve in fig. 2 is explained equally well as by the supposition that the positive ions are able to ionise the gas when theii velocity attains this value .
Experiments were therefore made to test these views as to the origin of the increased ionisation obtained .
Experiments at Different Pressures .
The pressure of the gas in the apparatus during the experiment , the results of which are recorded in fig. 2 , was ( M)69 mm. The mean free path of an electron in helium at this pressure is about l-7 cm .
The shortest distance between the glowing filament and the gauze was 3 mm. , but the periphery of the gauze was about 2 cm .
from the middle of the filament , so that electrons from the further parts of the gauze would mostly collide with gas atoms before reaching the filament .
The photo-electric emission of electrons from the gauze is , however , strongest at the parts of the gauze nearest to the filament , and here the path they travel through the gas is much less than their mean free path , so that the possibility of these electrons colliding with the gas atoms will be increased by raising the pressure until the mean free path of the electron is about 3 mm. If , then , the observed break in the curve at 20 volts is due to a radiation produced below the gauze by collisions of electrons with gas atoms , the effect should be more marked at a pressure of 0'3 mm. , at which the mean free path of the electron is 3 mm. , unless the radiation produced is very easily absorbed in the gas , in which case the increase in the radiation produced by the larger number of electron impacts may be neutralised by absorption of the radiation at the higher pressure .
In order to obtain an estimate of the mean free path of the positive ions from the glowing filament , it is necessary to make an assumption as to their nature .
If we assume that the mean free path of the positive ion in helium is about equal to that of a helium atom , the mean free path at a pressure of 0-069 mm. is about 3 mm. The positive ions from the glowing filament which take the shortest path towards the gauze will thus mostly reach the gauze without colliding with gas atoms ; but if the pressure is increased , many more of the ions will collide before reaching the gauze , and if ionisation occurs , the resulting electrons will go towards the filament and will not affect the collecting electrode in the ionisation chamber .
Some of the positive ions formed by ionisation below the gauze will pass through the gauze , but the velocities of these , and of those positive ions which have collided below the gauze , will be less than the velocity of the positive ions from the filament which reach the gauze without colliding , and they will be ) An Investigation of the Ionising Power Positive Ions .
341 less capable of producing ionisation on the other side .
If , therefore , the observed increase of current in the ionisation chamber which occurs when the difference of potential between the filament and the gauze reaches 20 volts is due to the positive ions which pass through the gauze , it should be less marked , on this account , at a higher gas pressure than it is at a pressure of 0-069 mm. This is also true if the positive ion is of larger dimensions than the helium atom , if , for instance , it is an atom of sodium or of potassium , as has been suggested by Richardson.* We are thus led to expect that at a pressure slightly higher than 0-069 mm. the increase of current measured by the electrometer when the potential difference accelerating the positive ions is increased beyond 20 volts , may become more marked if the effect is due to the production of radiation by electron collisions below the gauze , but should be less marked if it is due to the positive ions which enter the ionisation chamber .
It is , however , not so easy to distinguish between the two possible causes of the increase in the current with increasing accelerating potential as it appears to be from what is written above , because the magnitude of the current measured by the electrometer , due to the liberation of a given number of electrons inside the ionisation chamber , either by photo-electric action on the walls , or by collisions of the positive ions with the helium atoms , or with the walls , is not the same at different pressures .
At a pressure of 03 mm. electrons liberated from the walls of the ionisation chamber would be able to produce fresh ions in the gas by collisions if they were emitted from places near to the collecting electrode where the electric field is strongest .
If the gas pressure were reduced the number of new electrons formed from the gas by the collisions of electrons with helium atoms would probably increase at first , and then , on further reduction of the pressure , would decrease until finally , when the pressure was so low that the mean free path of the electrons became longer than the greatest distance of the collecting-electrode from the active parts of the chamber walls , the electrons coming from the walls would produce very few new ones by collisions with the helium atoms .
We have seen that if the rise in the curve ( fig. 2 ) at 20 volts is due to radiation produced below the gauze by collisions of electrons with gas atoms , the amount of radiation produced should be increased by raising the pressure from 0-069 mm. to about 03 mm. As the difference of potential between the collecting disc and the gauze is 83 volts , this increase of pressure would also have the effect of increasing considerably the possibility of ionisation by electron collisions within the ionisation chamber thereby making the effect of the increased radiation still more marked\#151 ; * O. W. Richardson , 'Phil .
Mag. , ' VI , vol. 16 , p. 740(1908 ) .
I Prof. F. Horton and Miss A. C. Davies .
assuming that there is no appreciable increase in absorption of this radiation with the increased pressure of the gas .
We have also seen that if the rise in the curve ( fig. 2 ) is due to the positive ions which pass through the gauze , an increase of gas pressure above 0069 mm. will reduce their power of producing electrons inside the ionisation chamber ; but this reduction of the ionising power of the positive stream may be counterbalanced by the fact that at the higher pressures each newly liberated electron will be more capable of producing others by collisions with the helium atoms .
For these reasons it was anticipated that the results of experiments at different gas pressures might not be easy to interpret .
Curves representing the results of a series of experiments at different gas pressures are shown in fig. 3 .
The temperature of the filament was maintained constant throughout these experiments , and was measured by the 0 10 20 30 40 50 60 70 POTENTIAL DIFFERENCE Fig. 3 .
optical pyrometer as 1430 ' C. The platinum disc which served as the collecting electrode was 2 cm .
above the gauze , and there was a constant potential difference of 83 volts retarding the positive ions in their journey towards it .
Of the observations recorded in the curves , those at 0*236 mm. were taken first , and the pressure was then gradually reduced .
It will be seen that at each reduction of the pressure there was an increase in the slope of the curve .
When observations were taken , beginning at the lowest pressure , it was correspondingly found that the upward slope of the curve was decreased by an increase of the gas pressure .
These results , therefore , support the* view that the observed rise in the current measured by the electrometer is caused by the positive ions which enter the ionisation chamber , but the i An Investigation of the Ionising Power of the Positive Ions .
343 production of a radiation by the collisions of electrons with helium atoms below the gauze is not definitely disproved , for the effect of this might be small compared with the effect of the positive ions , or the radiation might be readily absorbed by the gas at the higher pressures * It should be mentioned that , as time went on , the positive emission from the filament gradually diminished , and the upward slope in the curves at 20 volts became less marked .
It could , however , be restored by raising the temperature of the filament , and towards the end of the research it became necessary to raise the temperature considerably in order to obtain a marked rise in the measured current with increasing accelerating potential .
This decrease of the effect with decreasing positive emission from the filament , at a constant temperature , is further evidence that the effect is mainly due to the action of the positive ions after passing into the ionisation chamber , and not to a radiation produced by electron collisions below the gauze .
At pressures higher than those represented in the curves at 0'5 mm. ) , no upward bend at 20 volts occurred , but the current decreased slightly as the potential was raised from 10 volts to 60 volts .
This gradual decrease in the current with increasing accelerating potential is probablyfdue to some of the positive ions recombining with electrons inside the ionisation chamber , for the loss of electrons due to this cause would increase as the path of the positive ions through the gas was lengthened .
From fig. 3 it will be seen that the number of electrons emitted by the photo-electric action of the light from the glowing filament on the walls of the ionisation chamber varied when the gas pressure was changed .
The curves given show an increase in the photo-electric effect with decreasing pressures , especially at low pressures , but in other cases ( one of which is recorded in a later part of this paper ) , a small increase of pressure produced an increased emission .
These changes in the photo-electric effect with alteration of pressure at low pressures need further investigation .
Experiments in which the Field Retarding the Positive Ions was Varied .
It has already been pointed out that if the observed increase in the current between the gauze and the collecting electrode is due to the positive ions which enter the ionisation chamber , it must be caused either by the impact of these ions with the helium atoms or by their impact with the gauze or with the walls of the cylinder to which they are driven by the retarding * The authors have recently found that a radiation is produced when elections bombard helium atoms with a velocity of about 20 volts , but this effect was probably too small to have any marked influence on the results of the experiments described in this paper .
Prof. F. Horton and Miss A. C. Davies .
field .
It seemed possible to decide which of these is the cause by observing the increase in the current for a given increase in the difference of potential accelerating the positive ions , when these are subjected to different retarding fields after passing through the gauze .
Observations of the increase of the current were therefore made when the accelerating potential was raised from 20 volts to 65 volts , from 40 volts to 126 volts , and also through other ranges , in each case with two or three different values of the retarding potential .
The following table contains the values of the currents for 20 volts and 65 volts accelerating potential , with retarding potentials of 83 volts and 124 volts respectively .
The apparatus contained pure helium at a pressure of 0*00865 mm. and the distance of the collecting electrode above the gauze was 2 cm .
Potential difference accelerating the positive ions from filament to gauze .
Currents ( 1 = 3 x 10 13 ampere ) measured with different retarding potentials .
Retarding P.D. = 83 volts .
Retarding PJ ) .
= 124 volts .
20 volts 5650 5950 65 volts , # 6770 7700 Increase in measured current = 1120 1750 It will be seen that , although the chance of a positive ion colliding with a helium atom inside the ionisation chamber is greater with the smaller retarding potential , the increase in the measured current was actually smaller when the retarding potential was 83 volts than when it was 124 volts .
Clearly , therefore , with the accelerating potentials used the positive ions do not ionise the gas , and the increase in the current obtained on raising the accelerating potential from 20 volts to 65 volts must be due to the positive ions liberating electrons from the walls of the ionisation chamber .
On this view , the fact that the increase in the measured current was greater with the greater retarding potential is an indication that , although the positive ions do not ionise the gas , they lose energy as a result of collisions with the helium atoms , so that the energy of the bombardment of the walls of the ionisation chamber is increased when the path of the positive ions is shortened by increasing the retarding potential .
The pressure of helium used in these experiments has to be carefully chosen , for while it is necessary to have the pressure sufficiently high for most of the positive ions to make collisions with the helium atoms after passing into the ionisation chamber , it is also necessary for the pressure to be low enough for the mean free path of an electron to be great compared f An Investigation of the Ionising Power Positive Ions .
345* with the distance between the collecting electrode and the gauze , so that the-increase in the current in the ionisation chamber produced by the liberation of a given number of electrons from the walls shall not vary with the different retarding potentials used .
Several sets of observations were made at gas pressures slightly higher than those used in the experiments recorded above , and with accelerating potentials up to 126 volts .
With the larger increase of the accelerating potential a larger increase in the measured current was obtained , but the results of all the experiments agreed in showing that neither radiation nor ionisation was produced by the collisions of the positive ions with the helium atoms under the conditions of the experiment .
The increase in the current to the collecting electrode which occurred when the accelerating potential was raised above 20 volts must , therefore , have been due to the increased velocity with which the positive ions bombarded the gauze and the sides of the ionisation chamber after being turned back by the retarding field .
It must be pointed out that in these experiments any recombination between the positive ions which pass through the gauze and the electrons emitted by photo-electric action from the walls of the ionisation chamber tends to diminish the effect which would be caused by ionisation of the gas by collisions , for the conditions of recombination would be best when the path of the positive ion is longest , which is also the condition most favourable for ionisation .
It is , however , very unlikely that an increase in the ionisation with increasing path of the positive ions would be entirely masked by an equal increase in recombination , especially as we have direct evidence that the effect of recombination is small .
The curves in fig. 3 show an upward slope with increasing accelerating potential ; it follows , therefore , that more-electrons are produced by the positive ions which pass through the gauze than are lost by recombination in consequence of their increased paths .
The decrease in the upward slope of these curves as the gas pressure is increased may be accounted for on the supposition that the positive ions lose energy by collisions with the helium atoms .
These collisions will be more frequent at the higher gas pressures , and consequently , the energy with which the-positive ions bombard the walls of the ionisation chamber when a given accelerating field is used will decrease with increasing pressure , and from this decrease in the energy of the bombardment a smaller liberation of electrons results .
346 Prof. F. Horton and Miss A. C. Davies .
Experiments in which the Distance between the collecting Electrode and the Gauze was varied .
It is obvious that the best way to decide whether the increase in the measured current which occurs when the positive ions from the glowing filament are subjected to an increasing accelerating force is due to collisions of these ions with helium atoms , or with the negatively charged walls of the ionisation chamber , is to make observations when no gas is present , for then any effect obtained must be due to the latter cause .
The apparatus was , therefore , pumped out as completely as possible by means of the mercury vapour pump .
The charcoal tube was reheated , so that any gas which had been absorbed was driven through the apparatus , thus removing the last traces of helium .
The charcoal tube was finally cooled in liquid air for some time before the observations were begun .
During the experiments the pump was kept running continuously , and the pressure measured on the McLeod gauge was 0*00002 mm. The negative current to the collecting electrode was measured for several positions of this electrode above the gauze , in order to see if the current varied when the positive ions getting through the gauze travelled over different distances .
Observations were generally taken with three different accelerating potentials , e.g. , 42 volts , 126 volts , and 210 volts , and with a constant retarding potential which was usually about 30 volts greater than the largest accelerating potential to be used .
The numerical values of the currents measured by the electrometer were plotted against the corresponding values of the accelerating potential , and the points for each position of the electrode usually fell well on a straight line .
The straight lines for the different positions were nearly coincident , except in those cases where observations were made with the collecting electrode nearer than 1 cm .
to the gauze , in which case the currents measured were somewhat smaller than at longer distances .
After a set of observations had been taken in a high vacuum , the mercury pump was stopped and the stopcock to the helium reservoir was opened , thus allowing helium to enter the apparatus very slowly through the charcoal tube , which was cooled in liquid air .
When the pressure had risen sufficiently the temperature of the filament was adjusted by means of the optical pyrometer to the same value as before , and the series of observations was repeated .
In this way the effect of the presence of pure helium gas in the apparatus was studied .
As some of the curves obtained under different , conditions in these experiments very nearly coincide , the results are more clearly seen if presented in tabular form .
In the following tables , therefore , ) An Investigation of the Ionising Power of the Positive Ions .
347 we give the currents , measured by the electrometer , in a high vacuum and in helium , at a pressure of about 0-02 mm. of mercury .
The numbers given represent the currents with accelerating potentials of 42 volts and 210 volts respectively , at each of the distances noted .
It should be mentioned that , as usual , the currents were measured both when the potential difference accelerating the positive ions was being increased and when it was being decreased .
The values at a given voltage under these conditions were always very nearly the same and were often identical , showing that the positive emission had remained constant .
The numbers given in the table are the means of the values for increasing and decreasing accelerating fields .
Gras pressure in mm. Potential difference accelerating the positive ions from filament to gauze .
Currents ( 1=6 *8 x 10"12 ampere ) measured with different distances ( d ) between the gauze and the collecting disc .
d \#151 ; 0'o cm .
^=1*5 cm .
d = 2 *5 cm .
d = 4\gt ; *5 cm .
0 *00002 42 volts 210 volts Increase in measured current = 85 130 91 156 96 156 94 141 45 65 60 47 0 *0194 increasing to 0 *0202 42 volts 210 volts Increase in measured current = \#151 ; 144 249 134 224 113 166 \#151 ; 105 90 53 The temperature of the filament during these observations , as measured by the optical pyrometer , was 1550 ' C. , and a constant retarding potential of 248 volts was maintained between the gauze and the collecting electrode .
It will be seen from the table that the current measured by the electrometer when the accelerating potential was 210 volts , was considerably larger than that measured with an accelerating potential of 42 volts , even when the vacuum was as perfect as it could be made .
At the low pressure recorded , the mean free path of a helium atom is of the order of 10 metres , so that very few collisions between the positive ions and helium atoms can occur .
In this experiment , therefore , the increase in the measured current must be due to the bombardment of the upper surface of the gauze and the walls of the ionisation chamber by the positive ions .
Since even when the collecting electrode is at the greatest distance from the gauze there is practically no gas to interfere with the free paths of the positive ions , the number of electrons liberated by the bombardment of the positive ions might be expected to be the same for the four values of d , and we might expect the currents measured with the same potential difference accelerating the positive vol. xcv.\#151 ; a. 2 F Prof. F. Horton and Miss A. C. Davies .
ions to be equal at all four distances .
It must be remembered , however , that the negatively charged gauze and parts of the walls of the platinum cylinder which are exposed to the radiation from the glowing filament , are emitting electrons which stream towards the collecting electrode , and the positive ions pass through this electron stream .
Evidence that under these conditions a certain amount of recombination between the positive ions and electrons takes place has already been given , and it appears also in the smaller negative current measured with the stronger accelerating field , when the collecting electrode is 4'5 cm .
above the gauze , as compared with the current measured when this distance is 1*5 cm .
or 2*5 cm .
The area of the walls of the platinum cylinder which is exposed to the light from the filament depends on the position of the collecting electrode ; the area exposed being smaller when the electrode is near to the gauze than when it is further away .
The electron current measured by the collecting electrode will therefore vary on this account as the position of the electrode is altered , and these variations in the current will be greater when the distance between the gauze and the electrode is small than when it is large , for in the latter case the parts of the cylinder exposed or shaded by the motion are much further away from the source of radiation , and the photo-electric effect from them may be very small .
Experiments showed that it was only when the distance between the gauze and the collecting electrode was less than about 15 cm .
that the current measured with a small accelerating potential , was appreciably altered by a small movement of the electrode .
The greater shading of the walls of the ionisation chamber from the light of the filament accounts for the currents measured , when the collecting disc was 05 cm .
above the gauze , being smaller than those which were measured at larger distances .
Turning now to the currents measured when the apparatus contained helium gas at a pressure of about 0*02 mm. , we see that these currents are considerably larger than those measured under similar conditions in a vacuum , and also that the currents measured with an accelerating potential of 42 volts are no longer the same for the distances 1*5 cm .
, 2*5 cm .
, and 4'5 cm .
, but gradually decrease as this distance is increased .
Similarly , the currents measured with an accelerating potential of 210 volts decrease with increasing distance between the collecting electrode and the gauze .
It follows , therefore , that the larger currents measured are not due to the positive ions ionising the gas by collisions with the helium atoms , for this , with the voltages and pressure used , would produce larger currents at the greater distances on account of the increased number of collisions which occur over the longer paths .
The fact that the currents are larger than under similar I An Investigation of the Ionising Power of the Positive Ions .
349 conditions in a vacuum is perhaps partly due to an increase in the photoelectric emission from the walls of the ionisation chamber\#151 ; an effect which has been mentioned in an earlier part of the paper\#151 ; but it is also due in part to the electrons ionising the gas by collisions .
At a pressure of 002 mm. of helium , the number of new electrons liberated by electron collisions must be small , for the mean free path of an electron is greater than the greatest distance between the gauze and the collecting electrode .
On this account , a small increase only , in the current would be expected as the electrode is gradually raised .
Actually the measured current decreases considerably as the distance between the gauze and the electrode is increased , and this decrease must be due to a diminution in the number of electrons emitted from the walls of the ionisation chamber .
These electrons are liberated partly by the photo-electric action of the light from the glowing filament , and partly by the bombardment of the positive ions .
The photo-electric liberation of electrons tends to increase as the collecting electrode is raised ( thus exposing a larger area of the walls to the action of the light ) , but the number of electrons so liberated does not alter appreciably when the electrode is raised beyond 1*5 cm .
The decrease in the measured current must therefore be caused by a diminution in the number of electrons which are liberated by the bombardment of the positive ions against the platinum walls of the ionisation chamber as the path of these ions through the gas is lengthened .
This is no doubt due to the positive ions losing energy by collisions with the helium atoms .
At a pressure of 002 mm. the mean free path of the positive ion\#151 ; if it has about the same dimensions as the helium atom\#151 ; is about 1 cm .
, so that the number of collisions made increases considerably as the collecting electrode is raised , and consequently the energy with which these positive ions bombard the upper surface of the gauze or the sides of the platinum cylinder is less at the larger than at the smaller distances .
Several sets of observations , with varying distances between the gauze and the collecting electrode , were made under different conditions of gas pressure and accelerating potential .
The results were similar to those shown in the table and indicate that , with accelerating potentials up to 210 volts , the positive ions do not ionise the gas by collisions , although they lose energy by colliding with the helium atoms , and consequently produce a smaller liberation of electrons when they bombard the metal .
Direct evidence that the positive ions lose energy as a result of collisions with the helium atoms was obtained in experiments which were made to investigate the distribution of velocities among them as they pass through the gauze under a given accelerating force .
2 f 2 i Prof. F. Horton and Miss A. C. Davies .
The Distribution of Velocities among the Positive Ions .
The method which was employed by Pawlow and by Bahr and Frank for determining the distribution of velocities among the positive ions consisted in accelerating these through the gauze by a constant potential difference , and gradually raising the retarding potential between the gauze and the collecting electrode from zero until it was sufficient to prevent any of the ions from reaching the electrode .
The disadvantage of varying the retarding potential has already been pointed out , and in the present experiments this potential was kept constant , and the accelerating potential was gradually raised until the positive ions had sufficient velocity to reach the collecting electrode against the retarding field .
The existence of the large photo-electric effect from the walls of the ionisation chamber under the influence of the light from the filament makes exact measurements of the distribution of velocities difficult , for , at the temperatures used in these experiments , there was always an excess of electrons over positive ions reaching the collecting plate .
In order to get a measure of the velocities with which the positive ions get through the gauze , it was necessary to lower the collecting electrode near to the gauze , so that the distance between the gauze and the electrode was only a small fraction of the mean free path of the positive ions , in order that their motion towards the plate might not be hindered by collisions with gas atoms as well as by the retarding potential .
Under these conditions it was found that , at low pressures , a fall in the current measured by the electrometer took place when the accelerating potential exceeded the retarding potential , showing that at this point the positive ions were reaching the collecting electrode .
When experiments were made with the distance between the gauze and the collecting electrode greater than the probable value of the mean free path of the positive ions , it was found that the accelerating potential had to be raised considerably above the retarding potential before any decided fall in the current measured by the electrometer took place , and the fall was more gradual under these conditions .
These results show that , at low pressures , some of the positive ions penetrated the gauze with a velocity about equal to the accelerating voltage applied , but that the energy of these was rapidly lost when the distance through which they travelled was large compaied with their mean free path .
Discussion of the Results of Earlier Experiments .
It has long been known that when a metal is bombarded by positive ions with high velocities it emits electrons ; the experiments described in the present paper show that electrons are also emitted when a platinum surface i An Investigation of the Ionising Power of the Positive Ions .
351 is bombarded by positive ions having a velocity of about 20 volts , and that the emission increases aCs the speed of the positive ions is raised .
It seems probable that this is the explanation of certain results which were obtained by one of us in earlier investigations of the emission of positive ions from platinum and from certain salts when heated * In these experiments a strip of platinum was heated electrically between two parallel platinum plates about 1 cm .
apart , and the positive emission from the heated strip was studied in air and in other gases at various pressures , both when the platinum strip was heated alone and when it was coated with certain salts .
Observations were made with the heated strip at a constant temperature and with a constant difference of potential between the strip and the platinum plates which formed the cathode , and it was found that the thermionic current measured usually decreased as the gas pressure was reduced from 10 mm. to about 1 mm. , after which it increased more or less suddenly .
The increase of current with diminishing pressure continued until the pressure had been reduced to a few hundredths of a millimetre .
It then generally decreased as the pressure was still further reduced .
The curve in fig. 4 , taken from the Fig. 4 .
paper referred to , illustrates the magnitude of the increase in the current at pressures below 1 mm. , in the case of a platinum strip at 1190 ' C. , and with a difference of potential of 160 volts between the electrodes .
A probable explanation of this result is that as the gas pressure is reduced , the mean free path of the positive ions becomes longer , and the energy with which they bombard the platinum plates increases .
At a certain stage electrons are liberated as the result of this bombardment , and these electrons are able to ionise the gas by collisions , so that a considerable rise in the current takes place .
As the pressure is still further reduced , the liberation of electrons increases , but as their mean free path also increases , the number of new ions produced by collisions passes through a maximum value and finally decreases with each diminution of the pressure , thus accounting for the fall in the current measured at the lowest pressures .
The effect of the positive ions in liberating electrons from the cathode * F. Horton , 'Roy .
Soc. Proc. , ' A , vol. 88 , p. 117 ( 1913 ) .
Prof. F. Horton and Miss A. C. Davies .
probably also accounts for the fact that even at very low pressures it is impossible to saturate the current between a glowing anode and a surrounding cathode .
Experiments made by several observers have shown that an approximate saturation only is obtained , and that the current rises continuously as the potential difference between the electrodes is increased .
This difficulty in saturating the current is accounted for if the positive ions liberate electrons by bombarding the cathode .
The results obtained by Bahr and Frank can be explained from the conclusions arrived at in the present research .
In their experiments the photo-electric effect of the light from the glowing filament upon the platinum gauze which formed part of the boundary of the ionisation chamber must , from the design of the apparatus , have been much smaller than the photoelectric effect in the experiments described in the present paper , but the current due to the small photo-electric effect must have increased as the potential difference accelerating the positive ions was increased , because , with the method they used , each increase in the accelerating potential caused an equal increase in the difference of potential between the collecting electrode and the gauze .
Ultimately , ionisation of the gas by electron collisions would take place , and the current would then rise more rapidly .
It seems possible that at some stage after this began the current became large enough to be detected by the electrometer , and that the accelerating potential at which the current was first detected was taken as the ionising potential for the positive ions in the gas .
The retarding potential used was of necessity greater than the accelerating potential , so that electrons liberated from the gauze would ionise the gas by collisions before the potential accelerating the positive ions had attained the value of the ionising potential for electrons .
This would account for the accelerating potential at which a current is first detected being less than the ionising voltage in the case of electrons .
Bahr and Frank came to no definite conclusions as to the existence of a minimum ionising voltage for positive ions , for they found that a current was first detected at an accelerating potential which was lower the greater the density of the positive stream .
This result is explained by the fact that the larger positive emission was obtained by raising the temperature of the filament , a process which would also cause an increase in the photo-electric emission .
With low voltages the photo-electric effect is probably the only source of the ionisation detected by the collecting electrode , but as the voltage is raised , a point is reached at which the positive ions are able to liberate electrons from the gauze , and this effect is superposed on the photo-electric effect , and no doubt accounts for the rapid rise in the current which was found to occur with larger accelerating potentials .
f An Investigation of the Ionising Power of the Positive Ions .
353 Some of the above remarks apply also to the experiments of Pawlow .
In his experiments , the existence of a source of electrons inside the ionisation chamber seems to be proved by the results of the investigation of the distribution of velocities among the positive ions .
These were accelerated into the ionisation chamber by a constant accelerating potential of 10 volts or less , and Pawlow came to the conclusion that even at very low pressures none of the positive ions attained a velocity corresponding to the applied potential .
The method of experimenting was to accelerate the positive ions with a constant field and to increase gradually the retarding field until no current was detected by the collecting electrode .
At the low pressures used , at which the mean free path of the positive ions was greater than the distance between the electrodes , it seems probable that some of the positive ions would have a velocity corresponding to the applied potential difference , and the fact that the current as indicated by the electroscope became zero before the retarding potential had reached this value , does not necessarily indicate that at this stage no positive ions were reaching the electrode , but only that the electrode was collecting equal numbers of electrons and positive ions .
Both the experiments of Pawlow and those of Bahr and Frank are open to the objection that if an electrode covered with sodium phosphate or impure aluminium phosphate ( the activity of which is due to the presence of sodium impurities ) is used , it is probably surrounded by an atmosphere of salt vapour at the high temperatures employed , and the results obtained are more likely to be due to ionisation of this vapour than to ionisation of the residual gas in the apparatus .
The possibility of volatile impurity contaminating the gas in the experiments described in the present paper was very small , for when the final observations were taken the filament had been heated daily for several weeks , and in many of the experiments a stream of pure helium was passing through the apparatus .
Summary .
The experiments described in this paper show that the positive ions from a glowing tantalum filament do not ionise helium when they collide with the gas atoms with velocities up to 200 volts , but that they liberate electrons from a negatively charged platinum surface when they bombard it with velocities greater than about 20 volts .

