351 On the Effect of Temperature on Ionisation .
By J. A. Crowther , M.A. , Mackinnon Student of the Royal Society , , Fellow of St. John 's College , Cambridge .
( Communicated by Prof. Sir J. J. Thomson , F.R.S. Received April 1 , \#151 ; Read April 29 , 1909 .
) The effect of temperature on the ionisation produced in a gas by Rontgen rays was first investigated by Perrin , * who , using air , concluded that the total ionisation in a gas was independent of the temperature if the pressure was kept constant .
McClung , f however , who repeated these experiments later with air , carbon dioxide , and hydrogen , found that the ionisation in a gas was independent of the temperature if the density of the gas is kept constant ; that is , if it is heated at constant volume .
Although no source of error could be indicated in Perrin 's work , there was little doubt that the later experiments of McClung were correct , and that between the limits of his experiments ( 15 ' C. to 272 ' C. ) , and for the gases used , the ionisation produced by Rontgen rays was independent of the temperature when the gas was kept at constant density .
It is well known that the ionisation produced by rays of given intensity in certain gases and vapours , for example methyl iodide , ethyl bromide , or carbon tetrachloride , is much greater than that in air , or carbon dioxide .
The present investigation was made to discover\#151 ; ( i ) If the effect of temperature on the ionisation produced in these gases and vapours was the same as for air .
( ii ) If cooling down air to a temperature near its condensation point caused any appreciable alteration in the ionisation produced in it by rays of given intensity .
As it is almost impossible to clean out completely a vessel which has once contained organic vapours , the second experiment was performed first .
In order to perform the experiments at low temperatures , the ionisation chamber had to be constructed so that it could easily be inserted in an ordinary Dewar , tube .
Its dimensions , therefore , were kept quite small .
At the same time it was necessary that the insulations should be sufficiently far from the liquid air to enable them to be kept practically at the temperature of the room .
Any considerable fall in temperature of the insulations invariably led to the deposition of moisture upon them , and a consequent * ' Annals de Chim .
et de Phys. , ' [ 7 ] , vol. 11 , p. 496 , 1897 .
+ McClung , 'Phil .
Mag. , ' [ 6 ] , vol. 7 , p. 81 , 1904 .
Mr. J. A. Crowther .
[ Apr. 1 , breakdown of their insulating properties .
The form finally decided upon is shown in the figure .
To Cells To Cells To Elect , The body of the ionisation chamber A is a brass tube 10 cm .
long and 3 cm .
in diameter .
The electrode q which is to be connected to the electroscope , is a small aluminium plate 2J x 1 cm .
, surrounded by an earthed guard ring and insulated with quartz .
The high-potential electrode is a larger aluminium plate , placed parallel to q , and about 1 cm .
from it .
The electrodes pass out through two brass tubes each 8 cm .
long .
M is a heavy brass plate which serves the double purpose of maintaining the insulations at an equable temperature , and of acting as a support to the can A. When A is in liquid air , M can be kept warm by small heating spirals , and during the heating experiments it can be kept cool by ice .
As the insulations are thus shielded from temperature changes , the junctions here can be made air-tight with sealing wax .
The Rontgen rays enter the chamber A by a circular hole a , 5 mm. in 1909 .
] On the Effect of Temperature on Ionisation .
353 .
diameter , closed with very thin copper foil .
As A is intended to be heated up , it is obviously impracticable to have a sealing wax joint at a. By substituting copper for aluminium as the material of the window , it is possible to make the joint air-tight by soldering .
The electrode q is sufficiently far from the window a for all the soft secondary rays emitted by the latter to be absorbed by the gas before reaching it ; on account of the thinness of the foil used the penetrating secondary rays were not sufficient in quantity to have any appreciable effect on the ionisation .
The vessel A is adjusted with respect to the X-ray bulb , so that the rays pass through the chamber without striking the electrodes .
This adjustment is made by the use of a fluorescent screen before soldering on the bottom of the vessel , and the absence of any soft secondary radiation is tested for by drawing the ionisation-pressure curve for the vessel .
In a temperature chamber intended for work with gases at constant density , it is necessary to ensure that practically the whole of the gas shall be at the same temperature ; otherwise , since the pressure is the same throughout , there will be a transference of gas from the hotter to the colder parts of the vessel , and the density of the gas between the electrodes will change .
For this reason the chamber A was cut off from the gauges when readings were being made .
For the same reason the tubes through which the electrodes pass out of the chamber were made as narrow as possible , about 4 mm. internal diameter .
In this way the volume of gas under experiment not actually within the heating furnace did not exceed 3 per cent , of the whole .
In order to eliminate the effects of any alteration in the intensity or hardness of the rays , and also in order to be able to detect with more certainty any small alteration in the ionisation in A , a " balance " method was used .
The standard chamber S was of the usual flat box type in which the case forms the high-potential electrode , the other electrode consisting of an aluminium leaf stretched over a wire ring .
The flat sides of the chamber are of aluminium foil to allow of the entrance of the beam of rays .
The insulation , as in the other chamber , was of quartz .
The two chambers were connected to the same electroscope , and their high-potential electrodes were charged , by small storage cells , to equal and opposite potentials .
The ionisation currents were , therefore , in opposite directions in the two chambers .
It is well known that the relative amounts of ionisation produced by Rontgen rays in different gases vary with the nature of the rays .
In order to avoid , as far as possible , any disturbance in the balance between the two \#166 ; chambers , due to alterations in the hardness of the rays , the standard Mr. J. A. Crowther .
[ Apr. 1chamber S was always filled with the gas which was under observation at the time in the chamber A. In making an experiment , the two chambers were filled with gas from the same source to some suitable pressure , and the tap connecting A to the reservoir was then closed .
The chamber A then contained a definite mass of gas , which remained the same throughout the experiment .
The pressure in S was then altered until the magnitude of the current through the two chambers , when the rays were passing , was the same , as shown by the leaf of the electroscope remaining stationary .
The temperature of A was then altered , and a balance again obtained .
With air in the chambers it was found quite possible to balance to within 3 per cent , of the total ionisation in the chamber ; with ethyl bromide and methyl iodide , which were subsequently employed , an even greater degree of accuracy could be obtained .
Quite a small change in the ionisation in A , due to alteration of temperature , should , therefore , have been perceptible .
Experiments on .
The two ionisation chambers were pumped out repeatedly by means of a water pump , and filled with air which had been freed from carbon dioxide and water vapour by passing first through tubes containing caustic potash and calcium chloride , then through a long tube filled with a mixture of phosphorous pentoxide and asbestos ( the asbestos being used to prevent the pentoxide from caking , and clogging the tube ) , and finally through a plug of glass wool .
The vessel A was closed , and a balance obtained by altering the pressure in S. A Dewar tube filled with liquid air was then brought underneath A and gradually raised until the level of the liquid air was just below the upper rim of the vessel .
Violent ebullition at first took place , but after a few minutes a steady state was reached .
On again turning on the Kbntgen rays it was found that the balance was quite undisturbed , the ionisation in A being apparently unaffected by cooling it down in liquid air .
A change of as much as 3 per cent , in the ionisation would have been quite appreciable , but no such change was discovered .
The experiment was repeated many times , but always with the same result .
In some of the experiments a capillary gauge of fine bore was connected to the vessel A so that the pressure in A could be read at any stage in the experiment .
When the liquid air was applied the pressure fell rapidly at first , but finally became stationary at about 470 mm. of mercury , indicating a mean temperature for the air in A of below \#151 ; 170 ' .
The conduction of heat along the metal tubes leading from the vessel was probably sufficient to prevent the air in it from falling quite to the temperature of the liquid air .
As mentioned before , the electrodes were kept 1909 .
] On the Effect of Temperature on Ionisation .
warm by small heating coils wound round the tubes .
The flow of heat along the tubes was therefore considerable .
Taken in conjunction with the experiments of McClung , these experiments may be regarded as showing that between the limits \#151 ; 170 ' C. and +273 ' C the ionisation produced by Rontgen rays in a given volume of air at constant density is independent of the temperature of the gas .
Experiments on Vapours .
The experiments on vapours were made not by lowering but by raising the temperature of the ionisation chamber A. For this purpose a small electric furnace was wound round the outside of A , a thermometer being inserted between the furnace and the walls of A to indicate the temperature .
A very convenient method of making an electric heater for a purpose such as this is to thread a German silver wire through an asbestos tube and wind the two evenly round the outside of the vessel .
A few layers of asbestos .cloth are wrapped round the outside , and the lower end of the can is covered with a packing of loose asbestos .
The upper end , with the window a , has , of course , to be left uncovered in order to admit the beam of Rontgen rays .
With this device there is no danger of the different turns of wire slipping , when they expand on heating , and so causing short circuits in the furnace ; while , on the other hand , when repairs or alterations have to be made in the chamber A the whole furnace can be taken to pieces and rewound in a few minutes .
Methyl iodide and ethyl bromide were selected for experiment , these two vapours differing widely in their behaviour under the action of Rontgen rays both from air and from each other .
It may be mentioned in passing that .at high temperatures both these substances pass readily through indiarubber .
The connection between the vessel A and the gauges was therefore made by clipping a slightly wider glass tube over the brass exit tube of A , the joint being rendered air-tight with sealing wax .
The whole of the apparatus for passing the vapours into the vessels and for measuring their pressure was \#166 ; of glass .
Several observations were made with each of the two chosen vapours .
They all presented precisely similar features .
A sample set of readings for \#171 ; ach substance is given in the appended table .
As a balance method was employed , the readings in the columns headed .
" Ionisation " give the pressure in S of ethyl bromide or methyl iodide , as the case may be , necessary to balance the ionisation current through the same .substance in A at the stated temperatures .
It will be noticed that in both series of observations given in the table , VOL. LXXXII.\#151 ; A. Mr. J. A. Crowther .
[ Apr. 1 , and in fact in every series made with these vapours , there is an initial rise in the ionisation in the chamber A on first raising the temperature .
At a temperature of about 70 ' this increase ceases , and the ionisation in A remains constant within the limits of experimental error up to the highest temperatures reached .
Table .
Ethyl bromide .
Pressure , 146 mm. Methyl iodide .
Pressure , 110 mm. Temperature .
Ionisation .
Temperature .
Ionisation .
14 ' 62 0 15 ' 61 -7 89 79 2 72 67 0 136 79-0 103 67-5 173 79 '0 146 67-0 184 79 '4 184 67*2 This initial increase in ionisation is not of the same magnitude in different series of experiments made with the same vapour .
It varies , in fact , with the length of time during which the vapour has been standing in the chamber before the readings commence .
Thus , in a set of readings in which ethyl bromide was admitted into the chamber A less than an hour before the readings began , the ionisation on first heating increased from 70 to 75 , remaining constant at the latter value when the temperature was still further increased .
In the experiments recorded in the table , the chamber had been filled with ethyl bromide vapour overnight , and the initial increase is .
proportionately much bigger .
These results , and the fact that from a temperature of 'about 70 ' upwards , the ionisation at constant density is independent of the temperature , show that this initial increase in ionisation is not a genuine alteration in the ionising power of the rays with alteration in temperature .
It must , in fact , be ascribed to the vapour which condenses on the walls of the chamber at ordinary temperatures being driven off as the temperature is raised , and so increasing the effective amount of gas in the chamber .
It is easy to show that very appreciable amounts of vapour are so condensed , even when the vapour is considerably removed from its saturation point .
Thus in one experiment , when the vessel was filled with ethyl bromide and allowed to stand , the pressure fell from 260 mm. of mercury to 253 mm. in the first half hour , and finally , after standing all night , to 214 mm. When methyl iodide vapour was similarly left to stand in the vessel overnight the pressure fell from 215 mm. of mercury to 181 mm. .
1909 .
] On the Effect of Temperature on Ionisation .
357 It will be seen that these effects are of the same order of magnitude as the initial increase in ionisation on heating the chamber A. There can be little doubt , therefore , that this initial increase is due to the vapour being given off from the walls of the vessel as the temperature is raised .
We must conclude , therefore , that for the very ionisable vapours such as ethyl bromide and methyl iodide , as well as for air , the ionisation by Rontgen rays , when the density of the gas remains constant , is independent of the temperature of the gas .
Summary .
The ionisation produced by Rontgen rays has been measured in air at the temperature of liquid air , and in ethyl bromide and methyl iodide at various temperatures up to 184 ' C. It was found that in every case the amount of ionisation produced was independent of the temperature of the gas , if the density of the gas remained constant .
I have much pleasure , in conclusion , in expressing my best thanks to Prof. Sir J. J. Thomson for his constant and kindly interest throughout the course of the present experiments .
2 b 2

