]\gt ; On the Conductivity of Vapours .
By S. J. KALANDYK .
( Communicated by Prof. Sir J. J. Thomson , O.M. , F.R.S. Received June 5 , \mdash ; Read June 25 , 1914 .
) It has been shown by Beattie , *Garrett , and Garrett and Willows , when certain salts are heated up to 300-40 C. one can observe the discharge of positive and negative electricity .
This phenomenon is observed more particularly in the case of the halogen salts of cadmium , zinc and ammonium .
The researches of SchmidtS and Sheard have shown that the discharge is due to two factors , ( 1 ) emission of the charges from the surface of the heated salt , and ( 2 ) to the conductivity of the salt vapours themselves .
The thermionic effect of the above-mentioned salts was studied in detail by Prof. O. W. Richardsonl and his pupils , while some gations of the electrical properties of the salt vapours were made by Schmidt in the papers referred to above .
The object of the present research was to investigate the influence of the temperature , water vapour , time , etc. , on the current passing in the salt vapours .
The Thermionie Effect of the Salt and the onductivity of the Salt Vapours .
Some experiments were carried out in order to test whether the conductivity of vapours is due to charges which have diffused to other parts of the apparatus from the surface of the heated salt .
For this purpose the vapours were investigated after passing through a cylindrical denser in which the electric force was strong enough to sweep all charged particles to the electrodes .
The apparatus used consisted of a small bulb ( fig. 1 ) , in which the cadmium iodide was placed .
The vapour at first passed through the plug of glass wool to stop the large ions which might be produced during the heating of salt , then entered the condenser , where the smaller ions could be removed by the electric field .
The outer coating of the condenser was obtained by platinising the inner of the glass tube with a " " liquid platinum\ldquo ; ( from Deutsche Gold und Silber Scheide Anstalt ) and was connected to earth .
As the inner coating of the * Beattie , ' Phil. Mag vol. 48 , p. 97 ( 1899 ) .
Garrett , ' PhiL Mag ( 6 ) , vol. 13 , p. 728 ( 1907 ) .
Garrett and WiUows , ' PhiL Mag ( 6 ) , vol. 8 , p. 437 ( 1904 ) .
S Schmidt , ' .
Physik , ' 4 ) , vol. 35 , p. , and ( 4 ) , vol. 41 , p. 673.(1913 ) .
Sheard , ' Phil. Mag ( 6 ) , vol. 25 , p. 370 ( 1913 ) .
O. W. Richardson , ' Phil. Mag ( 6 ) , vol. 26 , p. 452 ( 1913 ) .
On the Conducti ity of Vapours .
condenser a gIass rod covered with the same substance was used , and was connected by a platinum wire with the source of electromotive force .
For measuring the ionisation of the vapours a second cylindrical condenser was used .
It consisted of an outer earthed platinum tube about 10 cm .
long , in the axis of which a short platinum tube was fixed , the latter being connected to a gold leaf electroscope .
Preliminary experiments were first made in order to test the action of the condenser .
For this purpose the bulb A was replaced by an open tube in which was placed a fine platinum strip which could be heated by an electric .
current .
The tube was connected with a water pump , and thus a stream of air was obtained which carried the ions emitted by the hot platinum upwards .
When the electroscope was charged up to volts , and both coatings of the condenser were earthed , it was found that the potential of the leaf fell 2 volts in five seconds .
It ; equired , however , about 90 seconds to fall the same amount when the inner surface of the condenser was charged up to volts .
It follows , therefore , that by this arrangement it is possible to stop practically all the ions entering the condenser To test if the ionisation observed in the salt Yapours is due to the surface effect , the bulb was filled with 5 .
of cadmium iodide , the whole apparatus was exhausted by cooled charcoal to the pressure mm. of mercury and heated in the electric furnacs to 29 C. When the current in the upper condenser constant the inner surface of the condenser was rged up to 200 volts as in the previous experiment , and the current in the condenser was measured during about 15 minutes .
The difference between the readings with and without charged did not exceed the error of the experiment .
'herefore we may conclude that the ionisation observed in the vapours is mostly due to the process which is going on in the vapours themselves , and is practically independent of the emission of charges from the surface of the heated salt .
It may be remarked that Sheard , working with apparatus of different construction , came to the same conclusion .
Mr. S. J. andyk .
Influence of Water * Willows , 'Phil .
Mag ( 6 ) , vol. 24 , p. 891 ( 1912 ) .
On the Conductivity of Salt were investigated in the tube ( 12 mm. diameter ) , which was connected through the tap with a manometer , pump and charcoal tube .
The salt was placed at the bottom of the tube and at a distance of 4 cm .
two platinum plates were fixed sq .
cm .
surface , distance between the plates cm One of the plates was connected to a Wilson electroscope divisions per volt ) and the second kept at the potential volts .
Preliminary experiments showed that , for the same potential difference , the current obtained was much smaller when the potential was applied for a long time than in the case when the electrode was charged only during the measurements .
Therefore , to keep the conditions steady , the plate was always connected with one pole of the storage cells .
The water was placed in a small bulb separated from the main apparatus by the tap L. ( The upper part was not used in this experiment .
) The lower end of the apparatus was placed in the electric furnace .
To free the bulb of air the water was frozen with liquid air and the whole apparatus exhausted with charcoal .
The stopcock was then turned on and the ice allowed to melt .
At first an experiment was made in order to test whether the presence of the water vapour had any action on the insulation .
For that purpose the empty apparatus was exhausted and heated up to about C. , then the stopcock was closed and opened .
observations showed that after about 15 minutes a small conductivity over the surface of the insulation appeared , but in any case this current was negligibly small in comparison with that which was obtained when the tube was filled with salt vapour .
The influence of water was tested for the salts cadmium iodide , zinc bromide , zinc iodide , potassium chloride , and potassium iodide , and the results obtained are set out in Table I , p. 638 .
In the case of cadmium iodide 1 .
was used and the apparatus exhausted to 10-4 mm. of mercury and heated up to C. In the case of zinc bromide 1 .
of the salt was placed in apparatus and kept one day in charcoal vacuum to liberate all water vapour , which is known to be very easily absorbed by this substance .
The pressure was mm. and the temperature 29 C. For zinc iodide .
was dried four days in a charcoal vacuum .
The temperature was 27 C. , and the pressure at the beginning of the observations was 3 .
10-6 mm. All the salts were obtained from Kahlbaum .
As may be seen from the Table I water vapour has a very marked influence on the current passing in the salt vapours .
The greatest increase of the conductivity was observed in the case of zinc iodide , which is also the most hygroscopic substance of those investigated .
The curious point is that the greatest conductivity is ob.served when the reservoir with the water is cut Mr. S. J. Kalandyk .
Table I. $ Arbitrary units .
off and the water vapour begins to get absorbed by the charcoal and phosphorus pentoxide .
This phenomenon may be due to the fact that the conductivity of the salt vapours depends not only on the presence of the water , but also on the pressure of the gas filling the apparatus .
It was shown by Schmidt that increase of pressure diminishes the current .
It may therefore be expected that the rapid absorption of water vapour will be accompanied by an increase of the current .
Of the salts which do not show any conductivity when heated , potassium chloride and potassium iodide were tested .
It was especially interesting to test the influence of water vapour on the last substance , as H. A. Wilson* observed a large current when a solution of I was sprayed into a hot platinum tube Garrett , however , noticed no leak when the dry salt was heated to the same temperature .
The experiments with this substance were carried out with a somewhat different arrangement .
The Wilson electroscope was replaced by a quadrant electrometer giving 50 divisions per volt .
The water vapour was absorbed by a tube immersed in solid carbon dioxide , and for measuring the current a small cylindrical condenser was used instead of the two platinum plates ; 3 .
of KI were placed the tube A and heated for two hours at a temperature of 30 C. The apparatus was exhausted with a Gaede pump to mm. * H. A. Wilson , ' Phil. Trans , vol. 197 , p. 416 ( 1901 ) .
On the Conductivity of Vapours .
The outer tube of the condenser was charged to volts , and the inner was connected with the eleotrometer .
At the commencement of heating a small leak was observed , which did not persist long , however , and the electrometer became perfectly steady .
When the stopcocks and were opened and a stream of water vapour was admitted into the tube A a considerable deflection was noticed ( five divisions in 10 seconds ) .
When the stopcock was closed the current disappeared also .
The phenomenon was repeated several times with identical results .
Potassium chloride was tested .
in the previous arrangement ( with the Wilson electroscope ) ; when the dry salt was heated to the temperature 30 C. at a pressure about mm. no current was observed as in the case of .
In the first moment when the water vapour was admitted the goId leaf showed a small and irregular movement ; this current , however , diminished gradually , and in a few minutes the water did not produce any leak .
Very likely the initial increase was also due to some secondary effect .
Change of Conductivity with the Time .
This question has been the subject many ations .
Garrett has shown that the conductivity produced by heated salts increases for about 20 minutes up to a certain maximum and then begins to fall .
The relation between the current and the time was found to be expressed by the formula where , and are constants .
his last paper Schmidt suggested the idea that the decrease of conductivity is due to two factors : ( 1 ) to the diminution of the surface of the salt during the heating , and ( 2 ) the change of concentration of ions due to their removal by the electric field .
The experiments described in the previous yraph show that the water vapour has enormous influence on the current passing in the salt vapours .
It was interesting , therefore , to test if the initial stage of conductivity is not also complicated by water vapour , which is usually absorbed by the salts and liberated during the heating .
For that purpose some measurements were made in conditions when all possible care was taken to get rid of traces of water vapour present in Jhe apparatus .
Cadmium iodide was used for this , as it does not contain any water of crystallisation , and therefore one can remove the water vapour absorbed by the salt by careful drying in vacuo .
The apparatus described in the previous paragraph was used for the measurements .
To have a supply of the dry air the bulb was replaced by a larger reservoir , in which the air could be kept in presence of phosphorus pentoxide .
To follow in more detail the initial stage of Mr. S. J. Kalandyk .
conductivity , the salt was not heated together with the tube , but it was placed in a small glass spoon , whence it could be thrown into the apparatus by turning the ground glass connection G. Two sets of measurements were made with cadmium iodide\mdash ; 1 .
The salt was placed into the spoon , the apparatus was exhausted to mm. and immediately heated .
2 .
The apparatus with cadmium iodide in the spoon was exhausted ith charcoal to about mm. and kept at that pressure for about five hours .
The apparatus was then filled with air which had remained a few days in contact with phosphorus pentoxide .
By opening the charcoal tube the pressure was then reduced to mm. The readings obtained in both cases iven in Table II and in the , fig. 3 .
Table II .
FIG. 3 .
On the Conductivity of Vapours .
The curve obtained in the first case ( fig. 3 , a ) does not differ from those obtained by the previous investigators if we take into account that the initial increase is due to the heating of the salt and to the diffusion of the vapours to the upper parts of the tube .
On the contrary , the current obtained with dry salt and dry air is practically constant .
It is interesting to note that , although cadmium iodide is not supposed to be very hygroscopic , nevertheless it liberates during the heating a considerable amount of water vapour , as will be seen from the change of pressure during the experiment .
In a fewminutes after throwing the salt into the tube A the pressure increased from to .
After that it began to decrease , and afterabout half-an-hour reached its initial value .
Further experiments with cadmium iodide were carried out in a high vacuum obtained by the cooled charcoal method .
It was expected that when the salt remained for a longer time in a very high vacuum and the gases liberated during the heating were instantly removed , the current passing through the salt vapours would be more constant .
For this experiment 1 .
of cadmium iodide was kept for one day in the highest vacuum obtained by charcoal , the apparatus was then heated .
and the salt thrown down .
The readings obtained are shown in Table III and fig. 4 .
During the last experiment the charcoal tube was open all the time and Table III .
FIG. 4 .
Mr. S. J. Kalandyk .
the pressure of air remaining in the apparatus was about 3 mm. The electrode was charged the whole time to volts .
The furnace was heated by the main battery of the laboratory and it was diffidult to keep the temperature of the apparatus more constant .
We may therefore regard the currents obtained after 20 minutes ' heating as practically constant .
Thus in the case of cadmium iodide it seems possible to obtain a fairly steady current in the vapours when the salt is properly dried .
Similar experiments were made with zinc iodide and zin6 bromide , but the influence of drying was not expressed so distinctly as in the previous case .
It may be that such a definite result could be obtained with cadmium iodide because this salt is not very easily volatilised and the change of conductivity due to the diminution of the surface of salt does not play a very great Influence of Temperature .
The influence of change of temperature on the conductivity produced by heated salts was tested by Garrett and by Schmidt .
In Garrett 's experiments the salt was pJaced on one of the places of the condenser aIJd hea , ted by an electric furnace surrounding the whole .
Thus , the current passing between the plates was due to the emission of the charges from the surface of the hot salt and to the conductivity of vapours .
The component of the current due to the conductivity of vapours was complicated by the change of pressure of the salt vapours .
Garrett observed that the current varied with the temperature according to Richardson 's law , and it therefore appears probable that in these conditions the thermionic effect of the salt plays the chief .
Further measurements of the temperature effect were made by Schmidt , who observed that the conductivity increases with temperature , at first slowly then more rapidly .
Schmidt 's arrangement allowed the of the vapours to be measured .
However , in his experiments also the constancy of the pressure was not taken into account and therefore the rapid increase of the current might have been due to the change of density of the vapours .
The following set of measurements was made in the conditions where the temperature of the vapours could be varied independently of the temperature of the salt .
Two grammes of cadmium iodide were placed on the bottom of a glass tube ( about cm .
diameter ) and heated in the electrical furnace to 20 C. At the distance of about 20 cm .
a cylindncal condenser was placed , consisting of a platinum tube 10 cm .
long and 2 cm .
in diameter , in the centre of which was fixed a narrow platinum tube 2 cm .
long and 2 mm. in diameter .
The outer electrode was to volts and the inner one connected to a Wilson electroscope .
The upper part of .
the tube was surrounded by a second On the Conductivity Vapours .
electrical furnace in such a way that the vapour passed a distance of about 15 cm .
in it before reaching the condenser so as to acquire the temperature of the second furnace .
The whole apparatus was exhausted and the salt was kept for one day in a charcoal vacuum .
During the experiment the pressure was 6 mm. In Table the measurements of the current at the various peratuoes are given .
It was found that the logarithms of the current are in a linear relation with the reciprocals of the absolute temperatures of the vapours ( see fig. 5 , A ) .
Therefore the connection between the current and the absolute temperature may be expressed in the following form , ( 1 ) where and are constants .
In the top column of Table the absolute temperatures calculated according to the formula are given .
The agreement with experiment is fairly satisfactory .
Table To follow the colmection between the current and the temperature over longer intervals some modifications were made in the previous arrangement .
The tube containing the salt was made of fused quartz ( diameter 1 cm .
) placed horizontally and joined to the apparatus by sealing wax cooled by a copper jacket .
For measuring the temperature of the furnace an iron-constantan -couple was used .
It was mentioned above that the electrodes placed in the vapours were connected continuously with the source of potential to secure greater constancy of .
As the increase of temperature produces a new concentration of ions it was necessary after obtaining the steady temperature to wait a certain time until the equilibrium was reached .
As this process goes fairly slowly it .
not possible to take readings at many temperatures , otherwise the conditions would not be identical .
The results obtained in this arrangement with , and are to be seen in the Table V or on the diagram , fig. 5 .
The discrepancy between the calculated and observed temperatures is somewhat greater in this case than in the previous one .
This disagreement may be due to the variation of pressure of the salt vapours during the experiment , VOL XC.\mdash ; A. 2 Mr. @f .
J. Table V. FIO .
6 .
which takes a considerable length of time ( about 3 hours ) .
Anyhow the linear connection between and 1/ 3 may be accepted as a first approximation .
The following formulae were used for calculation of absolute temperatures in the above Table:\mdash ; ; ZnBr : ; In the previous experiments it was shown that the conductivity in the salt vapours is mostly due to the processes which are going on in the vapours themselves .
The question now arises hither the observed conductivity is.due to the volume ionisation of the vapours or to the chemical action rs on the platinum electrodes .
An attempt was made to decide this point .
he tube , with a cadmium iodide , was On the lt Vapours .
placed horizontally , aild the current passing between the plates was measured .
One of the electrodes was then removed and covered with spongy platinum powder , so that its surface was greatly increased , while the volume of the vapour between the plates remained practically constant .
Therefore , if the observed ionisation was due to the surface effect , it might have been expected that the current would be much stronger than in the previous case .
However , the results showed that the current obtained with the spongy platinum electrode was of the same order as in the previous case , consequently this experiment seems to show that the conductivity is chiefly due to the volume ionisation .
If we assume that the conductivity of vapours is due to their electrical dissociation , we may easily calculate the work required to ionise one molecule of vapour , applying the formula of va n't where is number of calories required to dissociate one gramme-molecular weight of vapours , and are ratios between the number of dissociated and neutral molecules at the temperatures and T2 .
Calculating in this way the energy required to ionise one gratAme-molecular weight of vapours and dividing it by the number of coulombs required to electrolyse one gramme- equivalent weight of any substance , we obtain the values fO } the ionising potential of different salt vapours:\mdash ; ( olts ) .
Cadmium iodide . .
Zinc iodide . .
Zinc bromide . .
These values are much less than the corresponding potentials for gases .
Ionisation and oiation .
It is known that the salt vapours studied in the foregoing experiments are easily decomposed .
It seemed interesting , therefore , to test other stances the vapours of which are dissociated at high temperatures .
For that purpose phosphorus pentabromide and sulphur dichloride were tried .
The first compound is dissociated at C. and the second at C. The experiments were carried out in two different ements .
The first apparatus consisted of a tube with two platinum ectrodes , one of which was charged to 100 volts , and the second gonnected with a quadrant electrometer .
A small bead of phosphorus pentabromide was placed inside * H. A. Wilson , ' Phil. Trans , vol. 197 , p. 415 ( 1901 ) .
646 Mr. S. J. Kalandyk .
the tube , which was then exhausted with a Gaede pump and sealed off .
The tube was placed in an oven , which could be heated with a Bunsen FIG. 6 .
burner .
In the second arrangement the substance was placed in the small vessel ( fig. 6 ) , which was placed in an oven , the temperature of which could be regulated .
The vapours emerged through the tube , passed through the spiral , and came into the.tube , in which were placed two electrodes , one of them earthed , and the second connected with the electroscope , charged to about 300 volts .
The spiral and the tube were placed in the electric furnace , so that the temperature of the vapour could be varied without changing the temperature of the substance itself .
The experiments carried out with both arrangements have shown that the vapours of the above-mentioned substances do not conduct the electric current when they are dissociated by heat .
This negative result that the dissociation cannot be regarded as the cause of the ionisation .
Summary .
* The experiments described show\mdash ; 1 .
The conductivity of the salt vapours is due to the processes occurring in the vapours themselves .
2 .
The vapours of carefully dried salts conduct the electric current .
Therefore conductivity cannot be ascribed to the chemical action of water vapour on the salt vapours .
However , the presence of water vapour increases the current passing in salt vapours .
3 .
When cadmium iodide was very carefully dried , it was possible to observe a curren .
which was practically.independent of time .
* The preliminary account of this paper , containing the results 1 , 3 , and 6 , was read before a meeting of the Physico-Mathematical Society of Kiev , October , 1912 .
On the Conductivity of Salt Vapours .
647 4 .
The connection between the current and the temperature may be expressed with considerable accuracy by the formula where and are constants .
5 .
The ionising potential calculated from the energy of dissociation is considerably less than that for the ordinary gases .
6 .
The dissociation of vapours is not always accompanied by ionisation .
This research was carried out partly in the Cavendish Laboratory and partly in the physical laboratory of Kiev University .
I take much pleasure in thanking Sir J. J. Thomson and Prof. J. J. Kosonogov for having allowed me to work in their laboratories , and for the interest which they took in this research .
VOL XC.\mdash ; A. .

