408 Prof. F. Horton and Miss A. C. Davies .
mental tests , which are not chance samples of the mental elements , but are carefully chosen so as to measure different kinds of activity .
This application of Pearson and Filon 's theorem ( which contemplates , I take it , only sampling errors in the ordinary sense ) to the changes in correlation produced by sampling the underlying elements , is no doubt somewhat bold , and further investigation of an exact analytical nature will probably be fruitful .
Several difficulties remain , among which may be mentioned .the fact that in sampling the population in the case in question the correlations tend to err all in one way , either rising or falling , whereas here some seem to rise and some to fall , as though the " errors " are the departures , not from the equality point , but from some other origin .
In any case , the analogy between the experiments described and the theory that mental tests are samples of some kind , of mental elements is striking and suggestiveand there is evidently no need to postulate a general factor , as has frequently been done .
The hypothesis of a general factor is sufficient to explain the facts , but is not necessary , for a wider theory does so .
An Experimental Determination of the Ionisation Potential for Electrons in Helium .
By Frank Horton , Sc. D. , Professor of Physics in the University of London , and Ann Catherine Davies , M.Sc .
, Eoyal Holloway College , Englefield Green .
( Communicated by Mr. C. T. R. Wilson , F.R.S. Received December 21 , 1918 .
) Bohr has developed a theory of atomic structure by applying a system of mechanics based on the quantum theory of radiation to the theory of the nucleus atom proposed by Rutherford .
In the case of certain simple atoms he has calculated the minimum voltages through which an electron must fall in order to be able to produce ionisation by collision , * and a comparison of the calculated values with those obtained experimentally serves as a check on the validity of the theory .
Next to the hydrogen atom , that of helium has the simplest constitution , and this , for the normal atom , consists of two electrons rotating in non-radiating orbits round a doubly charged positive nucleus .
The application of Bohr 's formula to helium indicates that the minimum velocity necessary to remove one electron from the helium atom is about 29 volts , but the direct experimental determinations of several observers * N. Bohr .
, 'Phil .
Mag. , ' vol. 26 , p. 476 ( 1913 ) .
Ionisation Potential for Electrons m Helium .
40^ agree in fixing the value at about 20 volts .
Bohr has attempted to explain this by the assumption that the ionisation potential measured does not correspond to the complete removal of the electron from the atom , but only to a transition from the normal state of the atom to some other stationary state , where the one electron rotates outside the other .
When the outer electron falls back to its original orbit , radiation of sufficiently high frequency to liberate electrons by photo-electric action from the metal parts of the apparatus is produced , and Bohr has suggested that such a secondary effect has been mistaken for a genuine ionisation of helium by electron collisions .
The first direct determination of the minimum ionisation potential for helium was made by Frank and Hertz , * using the method originally devised by Lenard .
In their experiments , electrons from a glowing filament were accelerated towards a platinum gauze , on the other side of which they encountered an opposing field , which prevented any of them from reaching a collecting electrode .
If , however , the electrons on passing through the gauze possessed sufficient energy to ionise the gas present by collisions , the positive ions liberated were driven towards the collecting electrode , and the ionisation was measured by an electrometer in connection with it .
The experiment consisted in gradually increasing the potential difference accelerating the electrons until positive ions were detected by the electrometer .
The accelerating voltage at which this occurred was increased by the experimentally determined velocity of emission of the electrons from the glowing filament , and the value 20*5 volts was obtained for the ionisation potential .
A more recent determination by Lenard 's method was made by Pawlow , f who , as the result of his experiments , gave the ionisation potential of helium as 20 volts . .
A direct determination of the ionisation potential of helium by a different method has been made by BazzoniJ with an apparatus consisting of a hot tungsten cathode and a surrounding cylindrical anode .
Current-potential difference curves were plotted , and from the positions of the discontinuities in these curves , the value 20 volts was given as the minimum potential at which ionisation of helium by collisions takes place .
It is a curious fact that the values of the ionisation potential of helium which have been obtained by indirect experimental methods are more nearly in agreement with the theoretical value deduced from Bohr 's theory .
One-of the first of these was made by AstonS from observations of the length of * * * S * J- Frank and G. Hertz , ' Deutsch .
Phys. Ges .
Verh .
, ' vol. 15 , p. 34 ( 1913 ) .
t W. I. Pawlow , 'Roy .
Soc. Proc. , ' A , vol. 90 , p. 398 ( 1914 ) .
+ C. B. Bazzoni , ' Phil. Mag. , ' vol. 32 , p. 566 ( 1916 ) .
S F. W. Aston , ' Roy .
Soc. Proc. , ' A , vol. 80 , p. 45 ( 1907 ) .
Prof. F. Horton and Miss A. C. Davies .
the dark space which he discovered between the Crookes ' dark space and the cathode when a discharge is passed through helium .
By assuming that this represents the distance through which an electron travels from the eathode before attaining sufficient energy to ionise the gas by collision , Aston obtained the value 30 volts for the minimum ionisation potential .
C. and M. Cuthbertson have made measurements of the dispersion of many gases , and have found that their experimental results can be represented with accuracy by the formula 1 = C/ ( ?
io2\#151 ; n2 ) } where n0 and n are the frequencies of the free vibrations of the atom and of the incident light respectively .
They give , * as the value of n02 for helium , 3*499 x 1031 .
Using this value to calculate the ionisation potential by means of the quantum relation eY = Jin , taking e = 4*77 x 10"10E.S .
units and h = 6*55 x 10-27 grm. em.2/ sec. we obtain the value 24*4 volts .
The most recent indirect determination has been made by Richardson and Bazzonif from their investigation of the high frequency limits* of the spectrum of helium .
The terminal frequency observed in these experiments was identical with the frequency calculated from Bohr 's theoretical value of the ionisation potential using the quantum relation eY = hn .
In view of the remarkable agreement with experimental results of many of the deductions from Bohr 's theory , it seemed worth while to attempt another direct determination of the ionisation potential of helium , using a method by which a photo-electric effect of radiation could be distinguished from ionisation .
Description of the Apparatus .
Two sources of eleetrons were used in the course of these experiments ; the first of these is represented in position in the apparatus in the diagrammatic view in fig. 1 .
It consists of a tantalum filament , F , bent into the form of an inverted V , upon the apex of which rests a circular sheet of thin platinum foil , E , making contact at its centre with the filament .
The platinum foil was 0*01 mm. thick and free from pin holes , and was bound over the top of a short piece of glass tubing , cut from a ^-inch test-tube , which just fitted into the glass tube at the lower end of the apparatus , as shown in the diagram .
The wire binding the platinum foil in position was connected to a stouter platinum wire sealed through the glass at P. Before being fitted in position , a small area in the middle of the platinum foil was covered with lime by igniting on its surface a thin paste made by mixing finely powdered lime and sealing wax with water .
In the experiments the lime-coated * C. and M. Cuthbertson , ' Roy .
Soc. Proc./ A , vol. 84 , p. 13 ( 1910 ) .
t O. W. Richardson and C. B. Bazzoni , 4 Phil. Mag./ vol. 34 , p. 285 ( 1917 ) .
) Ionisation Potential for Electrons in Helium .
411 platinum foil was heated at a central spot by the tantalum filament in contact with it , the current to heat the filament being led into the apparatus through the stout platinum leads G and H. This arrangement gave an equipotential source of electrons , but it proved troublesome in practice because , after it had been used for some time , it became necessary to raise the filament to a high temperature in order to maintain a sufficient supply of electrons .
This sometimes led to the filament breaking through the thin platinum foil , and caused considerable delay in carrying out the research .
For this reason , in the later experiments , the platinum foil was discarded , and the bare tantalum filament used as the source of electrons .
This has the disadvantages , as compared with the former method , that it exposes the other metal parts of the apparatus to a good deal of light , and also that the electrons are emitted from parts of the wire at different potentials , which means that they will have different velocities when they enter the ionisation chamber .
This is , however , not a very serious objection , for the fall of potential down the glowing part of the filament used was probably never more than about 0*4 volt* and would cause a smaller range of velocities than that due to the Maxwell distribution of velocities among the emitted electrons .
On the other hand , the use of the tantalum filament alone enabled it to be heated to a white heat during the process of ridding the electrodes of occluded gases , and this process could consequently be carried out much more completely than when the equipotential source of electrons was ] used , so that with the bare tantalum wire there was less chance of the gas experimented on being contaminated during the experiments .
The battery supplying the accelerating potential difference was connected to the equipotential source of electrons at P , but in the experiments when E was removed and the glowing tantalum filament supplied the electrons , the potential was applied to the negative terminal H. Fig. 1 .
Prof. F. Horton and Miss A. C. Davies .
A circular piece of fine platinum gauze , D , was fixed 1*5 mm. above the filament , and a second similar gauze , C , 2 mm. above D and parallel to it , as shown in the diagram ; each of these was connected to an electrode sealed through the glass .
A third parallel gauze , B , was stretched tightly across the lower end of an aluminium cylinder , K , 3*5 cm .
in diameter , which just fitted into the upper part of the enclosing glass vessel .
The narrowing in of this vessel served to support the base of the cylinder symmetrically above the gauze , C , and at a distance of 1*2 cm .
from it .
Concentrically within the cylinder , K , a circular platinum plate , A , was suspended .
This acted as the collecting electrode and could be raised or lowered within the cylinder over a range of several centimetres .
This motion was secured by having the supporting silk thread wound round the plug of a glass stopcock , T. Between the plate , A , and the silk thread , a length of fine platinum wire was interposed , and this passed through , and made contact with , a small platinum ring at L , which was connected to the electrode , M , and thus to the sensitive electrometer used to measure the currents .
The inner surface of the aluminium cylinder was lined with thin platinum foil and its top was partially closed by a circular platinum plate , N , which had a central hole to permit the passage of the supporting wire .
It will thus be seen that all the exposed metal in the apparatus was platinum ; this was carefully cleaned by boiling for many hours in strong nitric acid before being fitted into the apparatus .
The glass containing vessel and the aluminium cylinder were carefully cleaned with chromic acid .to remove grease , and were finally washed out with distilled water .
The apparatus was connected , at the point indicated , through a glass stopcock with a wide bore , to a pressure gauge , a mercury vapour pump , and a long vertical tube 3 or 4 cm .
wide containing mercury , the level of which could be adjusted so as to vary the pressure of the gas in the apparatus .
Mercury vapour was prevented from passing over into the apparatus by a U-tube immersed in solid carbonic acid on the far side of the stopcock , and by a glass tube containing a plug of gold leaves and glass wool on the side of the tap nearer to the experimental tube .
The helium gas used in the experiments was purified by standing for several hours over charcoal cooled in liquid air , and was stored in a glass vessel , from which it could be allowed to pass into the apparatus by opening two stopcocks .
The gas passed very slowly along a fine capillary tube , and finally entered the apparatus below the gauze , B , through a U-tube containing charcoal and immersed in liquid air .
The final connection to the apparatus was 1*5 cm .
wide so that very free communication existed between the cooled charcoal and the ionisation chamber .
This is important beoause , Ionisation Potential for Electrons in Helium .
413 although some of the experiments were conducted with a stream of helium flowing through the apparatus , others had , of necessity , to be performed with a constant pressure of gas , and in these cases the purity of the helium was secured by the free connection with the cooled charcoal .
Before beginning the experiments , the apparatus was subjected to a long process of evacuation , the whole of the experimental tube , up to the level of the top of the aluminium cylinder , being enclosed in an asbestos box and heated to a temperature of about 300 ' C. During this process the mercury vapour pump was continually removing the gas evolved from the metal and the glass .
The filament was also heated to the limit of safety , and the heating and pumping were continued until the evolution of gas had practically ceased , a condition which it usually took about three days to attain .
Pure helium at a low pressure was then pumped through the apparatus for some hours before experiments were begun .
In order to prevent the electron stream from spreading laterally , the experiments were generally conducted with a magnetic field parallel to the axis of the tube .
This was obtained by wrapping many turns of wire round the apparatus and sending an electric current of 2 amperes through the coil while observations were being taken .
At one place a small gap was left between the turns of the coil through which the filament , F , could be viewed and its temperature was ascertained by means of a Cambridge optical pyrometer.* Experiments to Distinguish between Ionisation and Radiation .
With the apparatus illustrated in fig. 1 , three different methods of experimenting were employed .
The method used to distinguish between radiation and ionisation was similar to that used by Davis and G-oucherf in their experiments with mercury vapour and with hydrogen .
In this method the electrons from E were accelerated to the gauze , D , by maintaining a constant difference of potential ( Vi ) between these .
They were further accelerated from D to C by a potential difference ( V2 ) , which was zero to begin with and was gradually increased during the experiment .
The device of accelerating the electrons in two stages was adopted , because in preliminary experiments , made before the gauze , D , was introduced , it was found that the number of .
electrons passing through C increased with the increasing difference of potential between C and E , whereas it is desirable that the number should remain constant and only the speed vary .
It was found that * I am indebted to the Government Grant Committee of the Royal Society for the means of purchasing this instrument.\#151 ; F. H. t Davis and Goucher , ' Phys. Rev. , ' vol. 10 , p. 101 ( 1917 ) .
Prof. F. Horton and Miss A. C. Davies .
this condition was secured by interposing the extra gauze , D , and maintaining a constant difference of potential between D and the source of electrons .
A similar device was used by Pawlow in his experiments with positive ions .
The electrons which passed through the gauze , C , encountered an opposing electric field sufficiently strong to prevent any of them from reaching the gauze , B. The potential difference between C and B will be referred to as .
V3 , and this was usually maintained a few volts greater than the total accelerating voltage ( Yi + Y2 ) .
If ionisation of the gas occurs in the space* B C , some of the positive ions formed are driven by the field through the gauze , B , into the platinum cylinder .
If the direction of the electric field between A and B ( Y4 ) tends to accelerate these positive ions they will be collected by the electrode , A , and their presence indicated by the electrometer in connection with it .
If Y4 opposes the motion of the positive ions towards the collecting electrode , some of them may still reach it if V4 is small , and if the distance between A and B is small compared with the mean free path of the positive ions through the gas .
If radiation is produced by the collisions between the electrons and the gas atoms , this radiation will illuminate the plate , A , and also the inside walls of the platinum cylinder .
Under the influence of this radiation a negatively charged surface may be expected to emit electrons which will pass between the electrodes A and B , , and cause the electrometer to indicate a negative or a positive current according as B is negative or positive to A , that is , according to the direction of the field Y4 .
By keeping Y4 small and experimenting with it first in the ' same direction as Y3 , and then in the reverse direction , we are thus able to-distinguish between an ionisation effect and an effect due to radiation .
The curves shown in fig. 2 were obtained in this manner .
The upper curves , in which the current was always positive , were obtained with Y4 in the same direction as Y3 , and the lower curves with Y4 in the opposite direction to Y3 .
In the curves II and II ' the currents are plotted on ten times the scale used in the curves I and I ' , which represent a continuation of the same experiments to much larger values of the* accelerating potential .
From the curves II and II ' it is seen that an increasing current was detected by the electrometer when the potential difference accelerating the electrons reached about 21*5 volts , and that this current was reversed in direction by reversing the direction of V4 .
It follows , therefore , that it is due to radiation , and not to ionisation of the gas .
A careful comparison of the curves II and II ' and I and I ' shows that these are symmetrical up to 25 volts , but that after that velocity had been given to the electrons , the currents with V4 in the same direction as Y\amp ; ( curves I and II ) are larger than the corresponding currents with V4 in the Ionisation Potential for Electrons in Helium .
415 opposite direction to V3 ( curves V and II ' ) .
This indicates that at about 25 volts ionisation of the helium takes place .
In the experiments recorded in these curves the helium was at a pressure of O ' L45 mm. , V4 was 4 3 volts , and the distance between the plate , A , and the gauze , B , was 1 mm. It will be seen from curve I ' that , under these conditions , if ionisation of the gas occurs at about 25 volts , the positive ions formed are unable to travel from B to A against the retarding potential difference of 43 volts , until the + +s , I f n | j 1 i ' : ( h \#151 ; \lt ; \gt ; --\#169 ; \#169 ; j i S 1 \ \ 5 i \ 1 V1 A \ f \ i * / k \ K 0 10 20 30 40 50 60 70 Acce/ eraT/ ng pofenf/ s/ + / 2j vo/ ts Fig. 2 .
( Curves in lower part of this diagram should be I ' and IT not I Snd II .
) potential difference between C and B , which drives them through the gauze , B , has been raised to about 50 volts .
This is no doubt due to the positive ions colliding with the gas molecules and losing their velocity by these collisions .
It will also be seen from the figure that the curve takes an upward bend at an accelerating potential of about 55 volts , showing that , after this point , the increase in the number of positive ions collected more ' than counterbalances the increase in the number of electrons collected , a vol. xcv.\#151 ; a. o T Prof. F. Horton and Miss A. C. Davies .
state of affairs which continues with further increase of the accelerating potential , so that the measured current ultimately becomes positive .
In order to ascertain the actual velocity of the electrons corresponding to a given applied potential difference between E and C , the field between A and B was arranged so as to drive electrons from B to A. The " retarding field " ( V3 ) between B and C was then gradually diminished until electrons were just able to get through B , a point which was made evident by the negative charge gathering on the plate A. The difference between the retarding and the accelerating fields , that is V3\#151 ; ( V1 + V2 ) gives the correction to be added to the applied accelerating potential ( Vi + V2 ) , in order to obtain the velocity of the swiftest electrons passing through C. It is assumed that the point at which radiation or ionisation is first detected , is that at which the swiftest electrons are able to produce these effects , and the correction found , as described above , was therefore used in obtaining the minimum potentials at which the phenomena occur .
In the case of the results recorded in fig. 2 , the correction amounted to \#151 ; \#166 ; 11 volt , and the minimum radiating potential , as found from these experiments , is therefore 21*5 \#151 ; 11 = 20*4 volts .
The point at which the ionisation of the gas begins is not sufficiently sharply marked in the curves of fig. 2 to justify a deduction of the minimum ionisation potential being made .
It should be stated that in many cases , besides obtaining the maximum velocity of the electrons under the applied accelerating potential , curves were obtained from which the proportion of the electrons with any velocity between the maximum and zero could be deduced .
The form of these curves , in which the negative current measured by the electrometer is plotted against the retarding field , varied in different experiments .
Two typical curves are shown in fig. 3 .
Of these curve I was obtained , using the equipotential lime cathode source , and with pure helium streaming through the apparatus , so that the gas pressure rose from 0*074 mm. to 0*078 mm. during the observations .
Curve II was obtained with the tantalum filament as the source , of electrons , and with a pressure of 0*018 mm. of helium .
It will be observed that a much greater homogeneity of velocity was obtained with the equipotential source of electrons than when the tantalum filament alone was used .
This is probably due to the difference in the form of the source of electrons in the two cases , rather than to the fact that one was an equipotential surface , whereas the other had a small fall of potential along it .
In the latter case , that of the filament , the electrons are emitted in all directions and at different distances from the gauze , D. Experiments in which the velocities were measured at different gauzes seemed to indicate that some of the electrons from the filament passed round the gauzes , C Current- Ionisation Potential for Electrons in Helium .
and D , instead of going through them , and so did not acquire the velocity corresponding to the applied accelerating potential .
In the case of curve I , the applied accelerating potential difference was 19*3 volts , and the maximum velocity of the electrons , as deduced from the curve , was 18*5 volts .
The correction to be applied to the accelerating potential when this is in the neighbourhood of 20 volts is thus \#151 ; 0*8 volt .
The value of the correction varied slightly in different experiments .
When the lime-covered platinum source was used it varied from about \#151 ; *0*5 volt to \#151 ; 1*3 volt .
With the tantalum filament , the correction to the applied potential difference was always positive , and varied in different experiments from about 1*3 volt to 2-2 volts .
In the case of curve II of fig. 3 it was .
2*0 volts .
The corrections were found to remain fairly constant so long as the pressure of the gas and the temperature of the source were kept the same , but each experiment from which a value of the minimum radiation or minimum ionisation potential was determined , was immediately followed by a determination of the maximum velocity of the electrons in the cathode stream when the applied accelerating potential was slightly less than the critical value .
0P.O .
retard ; ng f/ ecfrons(Yo/ ts)^ / ( A / y / i / / / " f Fig. 3 .
5 10 16 20 25 30 35 Mcce/ erat/ nqpofent/ a/ ( Y , + / \#163 ; ; vo/ ts Fig. 4 .
We have seen that the positive ions formed when ionisation of the helium occurs appear to lose their velocity when they collide with the gas atoms , and that in the experiments represented by curve I ' of fig. 2 they were unable to get across to the collecting electrode against the field V4 ( 4*3 volts ) until the field V3 , which accelerated them , had been raised considerably above the value it had when ionisation of the gas began .
In order to be able to measure the ionisation potential of the gas by this method , it is necessary , therefore , to arrange the conditions so that the positive ions can get across to the collecting electrode when they are first produced .
This is secured by 2 l 2 Prof. F. Horton and Miss A. C. Davies .
reducing the value of the field V4 , which opposes their motion towards the electrode , and by arranging the gas pressure so that the distance they have to travel is about equal to their mean free path .
It is important that the pressure is not reduced too much , for the mean free path of the electrons in the gas is about six times that of the positive ions , and at very low pressures few collisions between the electrons and helium atoms would occur , so that very little radiation or ionisation could be produced under these circumstances .
The distance the positive ions have to travel can be reduced by lowering the collecting electrode until it is close to the gauze B ( fig. 1 ) , but here , again , it is possible to overdo the adjustment , for with the field V4 opposing the field V3 the radiation produced from the helium atoms is made manifest by the photo-electric action it has upon the walls of the platinum cylinder , and the area of these walls exposed to the radiation is rapidly reduced when the electrode A is lowered very near to the gauze B. The continuous curve in fig. 4 represents the results of observations taken with a pressure of 00645 mm. of helium in the apparatus and with the collecting electrode 4 mm. above the gauze which forms the base of the hollow platinum cylinder .
In these experiments V4 was 1*5 volt , and in the opposite direction to V3 .
From the curve it will be seen that the positive ions began to be collected when the potential difference accelerating the electrons was 23*75 volts , and that when this potential had been raised to about 29 volts the measured current became positive , showing that more positive ions than electrons were reaching the collecting electrode .
The correction to the applied accelerating potential in these experiments was found to be 4-2*0 volts , and this gives 25*75 volts as the ionising potential of helium .
It should be pointed out that in these experiments , since V3 is always slightly greater than the accelerating potential , electrons liberated from the under side of the gauze B , either by the light from the glowing filament or by the action of the helium radiation , may acquire a velocity sufficient to ionise the gas by collisions before those from E can do so .
Whether this occurs or not depends upon the velocity of emission of the photo-electrons from the gauze , and also upon the sign and the magnitude of the correction to the applied accelerating potential .
Experiments showed that the ionisation produced by these photo-electrons was too small to mask the effect of the electrons from the source E. This is no doubt due to the number of photo-electrons produced being negligible in comparison with the thermionic emission from the glowing cathode .
The minimum potential at which radiation is produced from the helium i Ionisation Potential for Electrons in Helium .
419 atoms can also be obtained from the curve in fig. 4 .
As the accelerating potential is gradually increased , the small positive current which the electrometer indicated began to increase after about 8 volts had been applied .
The current then rose steadily until the accelerating potential reached 18*4 volts , after which it decreased and became negative .
It is shown below that the gradually increasing positive current beginning at about 8 volts is due to positive ions which are liberated from the gauze C ( fig. 1 ) , by the bombardment of the electron stream from E which is turned back in the space B C. These positive ions are accelerated by the field V8 , towards the gauze B , and at this low pressure acquire sufficient velocity to pass through to the collecting electrode against the retarding field V4 .
That the positive ions from C can do this indicates that when positive ions are formed by ionisation of the gas they will probably be detected at once .
The applied potential difference at which the rise in the curve of fig. 4 ceases ( 18*4 volts ) , is that at which radiation is produced from the gas , for this with V4 opposing V3 , causes photo-electrons from the walls of the cylinder to pass to the collecting electrode , thus at once reducing the measured positive current .
The correction to be applied to this potential difference was found by experiment to be -f 2*0 volts , which gives the value 20*4 volts as the minimum radiation potential of helium .
That the gradual rise in the positive current up to 18*4 volts accelerating potential is due to positive ions liberated from the electrode C , was proved by repeating the experiment after removing all the gas from the apparatus .
Observations were taken with the charcoal tube in liquid air and with the mercury vapour pump running continuously , so that the vacuum was as perfect as could be obtained .
The positive currents measured under these conditions are represented by the dotted curve of fig. 4 , wdiich is drawn on the same scale as the curve obtained with helium in the apparatus .
Up to the point at which radiation was detected in the latter case , the two curves practically coincide , showing that the gradual increase in the positive current which occurred up to this point is not due to the presence of the gas .
The only possible source of the positive current measured in a high vacuum would seem to be positive ions liberated from the platinum gauze C by the electron bombardment .
If the bombardment produced a radiation , this by its photo-electric action on the walls of the platinum cylinder would have caused a current in the opposite direction .
It should be possible to make positive ions , produced by the electron bombardment and passing into the platinum cylinder , give rise to an emission of electrons by turning them back so as to bombard the upper surface of the gauze B ( as illustrated in fig. 5 ) , and by increasing their velocity Prof. F. Horton and Miss A. C. Davies .
sufficiently.* On testing this , by making V4 greater than V3 , it was found that a small negative current was detected , which gradually increased as the speed of the positive ions was increased by raising V3 .
This experiment is thus further evidence that positive ions are emitted by the gauze C when this is bombarded by the stream of electrons from the glowing cathode .
In order to examine the production of radiation without the photo-electric A Path of'electrons representedthus ---\gt ; Path of Positive/ on ?
representedtfws -- -\gt ; Fig. 5 .
effect which it produces being disturbed by the effects of ionisation , the field V4 opposing the passage of positive ions to the collecting electrode was made sufficiently large to prevent any of the positive ions , which are driven through the gauze B by the field V3 , from reaching the electrode A until large values of the accelerating potential had been reached .
Many experiments of this nature were made , and in the course of these , the gas pressure , the distance of the collecting electrode from the gauze B , and the difference of potential , V4 , between these , were varied over wide limits , but the curves obtained showed no sudden bend between the point at which radiation was first detected ( the " minimum radiation potential " ) and that at which the positive ions produced by the ionisation of the gas began to get through to the collecting electrode .
The curves in fig. 6 represent the results of one of these experiments , in which the gas pressure was 0'078 mm. , the collecting electrode was 4 mm. from the gauze , and the difference of potential between these was 7*2 volts .
In the dotted curve the currents represented in the earlier part of the continuous curve are plotted on a larger scale .
As the potential accelerating the electrons is gradually raised , these are able to * The emission of electrons under these conditions is shown to occur in ' Boy .
Soc. Proc. , ' vol. 95 , pp. 333\#151 ; 353 ( 1919 ) .
i Ionisation Potential for Electrons in Helium .
produce radiation from the helium atoms over gradually increasing distances on both sides of the gauze 0 ( fig. 1 ) , and thus the photo-electric emission of electrons from the walls of the platinum cylinder increases with ( Vi + V2 ) ; but it will be observed that the curve in fig. 6 shows no sign of any sudden -- -\#169 ; 2V-0- k V k i !
\ X \ \ O 10 20 30 40 50 60 Acce/ eratinq potentia/ 1\#163 ; ) Fig. 6 .
bend when ( Vi + \2 ) passes about 235 volts as would be expected* if the ionisation , which we have already shown begins at that point , is accompanied by the emission of another type of radiation .
Experiment to Test for Ionisation at the Minimum Radiation Potential .
The experiment just described has shown that the ionisation , which has been found to occur at about 25-7 volts , is not accompanied by an emission of another type of radiation from the ionised atom .
In order to be sure that no ionisation accompanies the radiation which is produced at about 20-5 volts , it is necessary to make a more delicate test than that afforded by the experiments , the results of which are represented in the curves of figs. 2 and 4 .
In these curves the measured current is due to the difference between the numbers of positive ions and electrons collected , and the curves merely indicate that at about 205 volts the effect of the radiation is greater than the effect of any ionisation that may take place .
It is possible to make the test for the occurrence of ionisation at about 20-5 volts more delicate by arranging so as to reduce the effects of the radiation and to increase the sensitiveness of the test for the production of positive ions .
In the experiment to which * correction to the applied potential difference ( V , + V2 ) in the t experiments illustrated m fig. 6 , was 2-2 volts .
If the minimum ionisation potential is about 25*7 volts , ionisation begins when ( V^Vg ) = 23*5 volts .
Prof. F. Horton and Miss A. C. Davies .
fig. 4 refers , the gas pressure was 0*0645 mm. , the distance between the collecting electrode and the upper gauze was 4 mm. , and the potential difference ( " V"4 ) retarding positive ions in this space was 1*5 volt .
In order to make the effect of ionisation more marked , the gas pressure was raised to 0*10 mm. , and the temperature of the electron source was maintained the same , so that more collisions between electrons and helium atoms could occur under these conditions than in the case of the former experiment .
The distance between the collecting electrode and the gauze B ( fig. 1 ) was reduced from 4 mm. to 0*5 mm. , thus considerably reducing the area of the walls of the platinum cylinder exposed to the radiation , while at the same time increasing the possibility of the positive ions being detected .
The collection of positive ions was further facilitated by reducing the opposing potential difference ( V4 ) from 1*5 volt to 1*0 volt .
The results obtained under these conditions are represented in the curve in fig. 7 .
As in fig. 4 , a dotted 15 20 25 30 Acce/ erat/ ng potent/ a/ ( V , + V2 ) vo/ ts Fig. 7 .
curve is drawn to represent the increase in the current , which was measured under similar conditions in a high vacuum .
The tw^o curves fell too close together to be shown separately in the figure until after ( V1 + V2 ) had been increased to 23*6 volts , which is therefore the first point where positive ions due to the gas were detected .
These experiments thus show that no detectable amount of ionisation of the helium occurs at about 20*5 volts , the electron velocity at which radiation begins . .
A More Sensitive Method of Determining the First Critical Potential .
Lenard 's method of determining the ionisation potential of a gas , and the modification of that method described in the present paper , have the defect Ionisation Potential for Electrons in Helium . .
that for potentials slightly greater than that at which ionisation or radiation begins , these effects can only arise from collisions of electrons with the gas | atoms in a thin layer of the gas on either side of the gauze through which the electrons are accelerated into the ionisation chamber .
The thickness of the layer of gas , in which collisions resulting in ionisation or radiation are [ possible , increases as the accelerating potential is raised and the effects of radiation or ionisation gradually become more marked .
When the critical potential is first reached , however , it is only the few electrons which happen to collide with the helium atoms in the interstices of the gauze which are able to produce the effect being sought for .
The discontinuity in the current-potential curve will obviously be more marked if , when the electrons have acquired sufficient velocity to produce ionisation or radiation , they maintain that velocity over a considerable length of path , and thus have a greater chance of colliding with a helium atom before their energy is reduced by the retarding field .
This condition can be secured with the form of apparatus described in this paper , by connecting together the gauzes B and C and retarding the electrons between B and the collecting electrode A , so that the course of an electron making no collisions in its upward path is that indicated in fig. 8 .
With this arrangement the electrons maintain the velocity which they have at C over the distance C B unless collisions with gas atoms occur ; so that , when Vi-bV2 reaches the ciitical value , an electron will be able to produce ionisation or radiation if it collides with a gas atom in this space .
The results of an experiment made under these conditions are represented in the curve of fig. .9 .
In this case the retarding field V4 was 30 volts , and was maintained constant during the experiment , and the pressure of helium in the apparatus was 0*0186 mm. If this curve is compared with curve I of fig. 2 , in which the current scale is the same , it will be seen that the first critical potential is much more sharply indicated in fig. 9 .
The second upward bend in the curve ( at 28*6 volts ) marks the point at which ionisation of the gas began .
The curve then rises steeply until some of the electrons from the glowing cathode get across to the collecting electrode .
Ionisation by collisions occurs chiefly in the space between the gauzes B and C , and some of the positive ions formed pass through the gauze B into the field V4 , which C D a. v2 Fig. 8 .
Current + Prof. F. Horton and Miss A. C. Davies .
drives them to the electrode A. By separate experiments , the correction to , the applied accelerating potential was found to be +2*1 volts at both critical potentials .
Many more points were obtained between 18 and 19 volts accelerating potential than can be shown in fig. 9 , and when these were plotted on a larger scale than that of this figure , the break in the curve was found to occur at 18*3 volts .
Thus , in this experiment , the minimum radiation potential for helium is found to be 20*4 volts , and the minimum ionisation potential 25*7 volts .
Acceferat/ ng potent is/ CK + V2 ) w/ ts Aicce/ erating potent/ a/ ( %+ yo/ ts Fig. 9 .
Fig. 10 .
In order to distinguish between ionisation and radiation by reversing the field between the collecting electrode and the gauze B , and yet to retain the feature of this sensitive method , another gauze would have to be interposed a few millimetres below B. In the present experiments it seemed unnecessary to modify the apparatus for this purpose , as the production of radiation at about 20*4 volts , and ionisation at about 25*7 volts , had already been demonstrated by the experiments described in an earlier section of the paper .
Moreover , without any alteration of the apparatus , the sensitive method may be used to distinguish between ionisation and radiation by making experiments at different gas pressures .
In the experiment , the results of which are recorded in fig. 9 , the pressure of the helium was 0*0186 mm. , and at this pressure many of the i Ionisation Potential for Electrons in Helium .
electrons would traverse the space between the gauzes B and C without colliding with a gas atom .
It follows , therefore , that , if the gas pressure is increased , many more collisions resulting in radiation will occur at the minimum radiation potential , so that a rise in the current-potential difference curve , due to radiation , should be more marked at a higher gas pressure .
At the minimum ionisation potential the number of positive ions produced will similarly be increased by increasing the gas pressure , but , in the absence of an electric field to drive these positive ions towards the gauze B , fewer of them will pass through to the collecting electrode at the higher pressure , since not only is their rate of diffusion less , but fewer are formed within a short distance of this gauze .
Experiments were therefore made at several different pressures , higher than 0*0186 mm. , and these confirmed the results obtained by the method described earlier in this paper , namely , that radiation is produced from the helium atoms by impacts of electrons with about 20'4 volts velocity , and ionisation by impacts with about 25'7 volts velocity .
Fig. 10 represents the results of such an experiment with a pressure of 0*30 mm. of helium in the apparatus , and it will be seen that the break in the curve at the first critical potential is more sharply marked than in fig. 9 , but that there is , at this pressure , no sudden rise in the curve at about 23*6 volts .
The first critical potential is thus the minimum radiation potential , and the second critical potential is that at which ionisation first occurs .
It may be mentioned that by maintaining a small difference of potential between the gauzes B and C , so as to drive the positive ions formed upwards , the ionisation point may be made more marked at the higher pressures ; but for the purpose of obtaining the two critical potentials in the course of one experiment , the method was found to give the best results when the gas pressure was low and the gauzes B and C were connected together .
From the curve in fig. 10 it will be noticed that although the gas pressure was so high that every electron had probably made more than one collision with a helium atom in the space between the gauzes B and C , yet some of them traversed that space without appreciable loss of energy , and were detected by the electrometer when their velocity ( Yithe velocity of emission ) was equal to the retarding potential difference V4 ( 30 volts ) .
This must evidently mean that the atoms with which they collided had already been ionised by other electrons .
This effect did not occur when the temperature of the source of electrons was lowered so as to reduce the intensity of the electron stream , and in this case electrons were not driven through the retarding field Y4 until much higher accelerating potentials were used .
Prof. F. Horton and Miss A. C. Davies .
A Method of Determining the first Critical Potential , which does not involve a Correction for the Velocity of Emission of the Electrons from the glowing Cathode .
The methods hitherto described have required a separate experiment to determine the correction which must be added to the applied potential difference accelerating the electrons in order to obtain their exact velocity .
The question arises as to whether the few electrons which first pass through the gauze C ( fig. l ) , xwith the necessary velocity are able to produce an effect large enough to be detected by the electrometer , as we have assumed in deducing the minimum potentials at which radiation and ionisation occur .
The following method has the advantage that the critical potential is determined directly from the retarding field without the necessity of finding the correction to the applied accelerating potential .
The method depends on the fact that in helium the collisions of electrons with the gas atoms have been shown to be perfectly elastic when the electron velocity is below the critical value.* The pressure of the gas in the apparatus was adjusted so that the mean free path of the electrons was considerably less than the distance between the gauzes B and 0 , in order that it might be assumed that no electrons pass through this space without colliding with helium atoms .
The temperature of the glowing filament was maintained below the value at which the number of electrons became so great that some of them passed through the space between the gauzes B and C without hitting a normal helium atom .
Vi was as usual kept constant and equal to 7 volts , V2 was varied , V3 as in the method just described was zero , and the electrons were retarded by the field V4 , which , in the present method , was adjusted during the course of the experiment to the radiation potential .
It will be seen that with this arrangement , if V4 is made a little greater than the radiation potential , no electrons will reach the collecting electrode until the accelerating field ( Vi + V2 ) has been raised considerably above the radiation potential , for when the velocity of an electron becomes equal to the radiation potential , this velocity is lost on collision with a helium atom .
On the other hand , if V4 is made less than the radiation potential then , when V2 is gradually increased , the electrons will be able to pass to the collecting electrode as soon as their velocity becomes equal to V4 , and until it becomes equal to that corresponding to the radiation potential , when they will lose their velocity by colliding with the gas atoms in the space between the gauzes B and C. The method therefore consists in fixing V4 at a value in * J. Frank and G. Hertz , 6 Deutsch .
Phys. Ges .
Verb .
, ' vol. 15 , p. 34 ( 1913 ) ; J. M. Benade and K. T. Compton , 4 Phys. Rev. , ' vol. 11 , p. 184 ( 1918 ) .
Ionisation Potential for Electrons in Helium .
427 the neighbourhood of the radiation potential , and in determining whether this value is too low or too high by testing whether electrons can be driven across to the electrode A by gradually increasing the accelerating potential .
The lowest value of V4 at which no negative current can be detected by the electrometer , as V2 is gradually raised , gives the minimum velocity which the electrons must have in order to lose their energy on colliding with the helium atoms .
It was found that this point could easily be fixed to 1/ 10 volt , the smallest amount by which V4 could be varied with the arrangement used .
Observations were taken at a gas pressure of 0*87 mm. and with the electron source at various temperatures between 865 ' C. and 1050 ' C. , so that the intensity of the electron stream varied considerably .
The extreme values of the radiation potential measured in these experiments were 20*3 volts and 20*45 volts , and the mean value was 20*4 volts .
These experiments were performed using the tantalum filament as the source of electrons , and the velocity of the electrons was not sufficiently homogeneous for the method to be used to determine the minimum ionisation potential .
For this to be possible , the range of velocities among the electrons must be small compared with the difference between the ionisation and radiation potentials .
Under these conditions , if V4 is fixed at a value less than the difference between the ionisation potential and the radiation potential , and V2 is gradually raised , starting from a value for which the velocity of the electrons is slightly greater than that required to produce radiation , the electrons will get across to the collecting electrode as soon as their velocity exceeds the radiation potential plus V4 .
On the other hand , if V4 is fixed at a value greater than the difference between the ionisation potential and the radiation potential , electrons will not be detected by the electrometer until their velocity exceeds the ionisation potential plus V4 .
In this way the difference between the ionisation potential and the radiation potential may be determined , and , as the latter has already been found , the ionisation potential can be deduced .
Summary of Results and Conclusion .
The results to which reference has been made have shown that radiation occurs when electrons having a velocity of about ' 20*5 volts collide with helium atoms , and that this is not accompanied by any ionisation of the gas .
It has also been shown that ionisation of the helium occurs when the velocity of the electrons is raised to about 25*7 volts , and that no other type of radiation is produced at this point .
Accurate determinations of these critical potentials were made by making detailed investigations of the positions of the bends in the current-potential 428 Ionisation Potential for Electrons in Helium .
difference curves , by observations at intervals of 1/ 10 volt over a range of 2 or 3 volts , in the neighbourhood of the values already found .
In these experiments observations were taken , both when the accelerating potential was gradually increased and when it was gradually decreased over this range , and in the latter case the bends in the curves were often more sharply marked , and the values found for the critical potentials were sometimes a little lower , than when gradually increasing potentials were used .
The mean values found from these detailed experiments by the first method described were 20*4 volts and 25*7 volts respectively .
By the method in which V3 was made zero and the electrons could produce ionisation or radiation over longer paths , the mean values 20*45 volts and 25*5 volts were obtained .
The mean value of the potential at which radiation is first produced , found by the method which does not involve a correction for the velocity of emission of the electrons , has already been given as 20*4 volts .
We may therefore take 20*4 volts as the minimum radiation potential and 25*6 volts as the minimum ionisation potential of helium .
These experiments thus confirm Bohr 's suggestion that radiation and not ionisation occurs at about 20 volts , but the value we have found for the ionisation potential is considerably lower than the value\#151 ; 29 volts\#151 ; deduced from Bohr 's theory of the structure of the helium atom .
In an attempt to explain the results of Bazzoni 's experiments , in which ionisation of helium was apparently obtained at 20 volts , Van der Bijl* has adopted Bohr 's suggestion that radiation only is produced by 20-volt impacts , and has pointed out that the velocity of emission of electrons from the filament , under the influence of this high-frequency radiation , would be sufficient to give the photo-electrons a velocity greater than the calculated ionisation potential , so that the helium would be ionised by this means as soon as radiation appeared .
It has already been explained that in the present experiments the minimum ionisation potential was deduced by assuming that the critical point was due to the collisions of the swiftest electrons which could be detected coming from the filament , and the greatest velocity of emission which was observed was practically the same for accelerating potentials above and below the radiation potential .
This indicates that the low value found for the ionisation potential , as compared with the value predicted by Bohr 's theory , is not due to high speed photo-electrons coming from the filament .
There still remains the possibility that high-speed electrons liberated by * H. J. Van der Bijl , 4 Phys. Bev .
, ' vol. 10 , p. 546 ( 1917 ) .
On the Elasticity of Metals as Affected by Temperature .
429 the photo-electric action , on the under side of the gauze B , of the radiation from the helium atoms , may ionise the gas between the gauzes B and C. Such high-speed electrons , since they would be directed downwards , would not be detected in the velocity distribution curves .
That the ionisation produced at 25'6 volts is not due to this cause , was proved by making a series of experiments , using the method first described in this paper .
In the first experiment of this series , the excess of the retarding potential V8 over the accelerating potential ( Y1 + V2 ) was 1*4 volt .
In the succeeding experiments larger values of this excess were used up to a maximum of 32 volts , but the other conditions under which the experiments were conducted remained the same throughout the series .
With the larger values it is clear that photo-electrons from the gauze B would be able to ionise the gas when the accelerating potential was large enough to produce radiation from the helium atoms , but in no case was ionisation detected until the velocity of the electrons from the glowing cathode was raised to about 25'6 volts .
The ionisation which is suddenly detected at this point , is thus evidently not due to the action of photoelectrons produced by the 20'4-volt radiation .
Note on the Elasticity of Metals as Affected by Temperature .
By A. Mallock , F.R.S. ( Received January 1 , 1919 .
) When two solids or structures , identical in form and differing only in the material of which they are composed , are subjected to stress , their strengths depend on the limits of the strains which those particular materials will bear without permanent set or rupture , but their stiffnesses are directly proportional to the elastic constants .
It is with the strength of structures that engineers are mostly concerned , and a great deal of work has been done on the strength of materials for various classes of strains , of which some , though not much , relates to the variation of strength with temperature .
As regards the elastic constants themselves , however , there has been no systematic investigation of their temperature variation , though a few measures of the kind have been made in Switzerland and America .
The subject is not without interest both from a purely scientific and a practical point of view ; for , as will be seen from this note , there is a relation

