]\gt ; 1909 .
] Electricity through Gaseous as soon as the rays had passed , and the actual curve obtained would be of the form MPQrstST , in which the point corresponds to the receipt by A of all the positive ions of larger mobility , and the point to the receipt of all the more slowly moving ions .
Thus an experimental realisation of the curve expressing the relation between and should decide the question as to whether there are two distinct mobilities for the ions of any one sign in a mixture of gases .
Experimental Procedure .
The diagram of connections is represented in fig. 1 ; and are two iron weights , which are supported by means of an electromagnet ; on breaking the circuit the magnet these weights fall simultaneously and break the platinum contacts at and respectively .
The breaking of the contact at , which is in the primary circuit of a Marconi induction coil , ives rise to an induced E.M.F. in the secondary , and causes a momentary discharge to pass in the Rontgen bulb X ; when the contact at is broken the potential of the electrode of the ionisation chamber is reversed , as is manifest from the diagram , being a water resistauce of the order of 1 , and at the moment of breaking being at the potential of the point .
The time interval between the Bontgen ray flash and the subsequent reversal of potential could be adjusted by suitably fixing the position of with regard to a graduated vertical scale ; the actual values of the time interval were calculated by means of t , he ordinary formula for a body falling freely under gravity .
The bottom of the falling weight was at a height of cm .
above the contact K. is a capacity of about 7 microfarads inserted in parallel with the primary of the induction coil in order to prevent sparking at the contact ; when the spark was entirely eliminated , and when the current through the primary of the induction coil was kept constant , the intensity of the Rontgen ray flash was sensibly constant at each discharge .
A Dolezalek electrometer with a fine platinum suspension was employed in order to measure the ionisation produced ; with the needle charged to a potential of 80 volts , the electrometer afforded a deflection per volt of 620 mm. on a scale about 1 metre distant .
In the previous paper ( loc. cit. ) a brief discussion is given with regard to the theoretical assumptions and experimental difliculties in connection with Langevin 's method ; for a more detailed account the reader is referred to Langevin 's original memoir .
* A reading was taken in the following manner:\mdash ; ( i ) is adjusted to the required height on the scale .
* Ann. de Chim .
et de Phys VII , , p. 495 , 1903 .
1909 .
] Electricity through Gaseous Mixtures .
given in fig. 3 ; these correspond to positions of the slit between the lead plates , so that the layer of ions produced by the Bontgen ray flash did not graze the electrode ; the width of the slit between the plates was about 3 mm. It is noticeable that there is no evidence of the existence of a portion of the curve corresponding to the part in fig. 2 .
The curvature of the oblique portions is due probably to non-uniformity in the distribution of the ions .
In the present instance , the fall of potential across the electrodes was 242 volts , and the times taken for the positive and negative ions to traverse a distance equal to that ( 31 mm. ) between the two electrodes were and sec. respectively , these time intervals to the sum of the abscissae of the points and in the case of the positive ion and and in the case of the negative ion .
The resulting mobilities are cm . .
for the positive ion and cm . .
for the negative The curves in fig. 3 are typical of other curves which were obtained for the same mixture corresponding to different positions and different widths of the slit .
Ethyl Ether and Air.\mdash ; Fig. 4 represents similar curves obtained for a mixture of ethyl ether ( 122 mm. ) and air ( 508 mm the pressures 1909 .
] Electricity through Gaseous dioxide and passed through air was shown to have the same mobility as an ion formed in air and passed through air .
The experimental results to be described later lead , however , an entirely different explanation of the statistical average .
There is an interesting feature in connection with the curves and 4 which suggests at first sight the existence of an effect .
It is noticeable that the distance along the time axis corresponding to the points and is greater than the corresponding distance for the points and ; and similarly that the distance ] the time axis for the points and is greater than the corresponding distance for the points and ; in other words , the ions require a longer time after their arrival at A before they are all received by the electrode A than what is necessary for them to be all received by after their arrival .
It appears , therefore , that the layer of ions broadens out in its passage through the medium .
In order to test this point more completely , curves were drawn with oxygen alone in the ionisation chamber .
These curves are given in , from which it can be readily seen that the broadening out of the layer of ions occurs markedly for Bffect of Small Traces of Vapours on the Values of the Jfobilitips of the Ions form ed in Air .
It has been shown by Zeleny and by Rutherford that the vapours of water or alcohol when present in small quantities in air produce a marked diminution in the mobility of the negative ion .
There was very little effect on the value of the mobility of the positive ion in air ; in carbon dioxide the positive ion increased in mobility when water vapour was added and in hydrogen it decreased .
The following table exhibits the values of the positive and negative mobilities in the dry gases and also in gases saturated with water vapour\mdash ; 1909 .
] Electricity Gaseous JIixtures .
different from with a velocity which approximates to that which would ensue if the ionisation had been due to the hydrogen .
If we consider for the moment the positive ion only and if we regard it as the molecule of methyl iodide with which is associated a charge ( e ) equal to that carried by the monovalent ion in electrolysis , it is easy to calculate a superior limit to the velocity with which such an ion would move under unit electric intensity through hydrogen at C. and at a pressure of 1 atmosphere .
The formula given by the writer ( loc. cit. ) is where denotes the mass of the methyl iodide molecule , mass of the hydrogen molecule , radius of the methyl iodide molecule , radius of the hydrogen molecule , coefficient of viscosity of hydrogen , specific inductive capacity of hydrogen , pressure of hydrogen in c.g.s. units , density of hydrogen at 760 mm. pressure , mean velocity of agitation of a hydrogen molecule , A quantity The factor in the brackets corresponds to the diminution in mobility resulting from the inductive attraction between the ion and the neighbouring molecules .
The experimental values of the quantities involved in the calculation are In order to calculate it was taken as equal to , where and denote the densities in the liquid state of methyl iodide and hydrogen respectively .
With , and we have On making the necessary substitutions it is found that the mobility ( under 1 volt/ cm .
) of the methyl iodide molecule through hydrogen is if we the diminution in the mean free path due to the attraction of the molecules , while if this attraction be taken into account the value of is .
Of course , any clustering round this molecule as nucleus would produce a further diminution in the value of the mobility .
The calculated mobility values , even under circumstances tending to increase the velocity , are thus considerably less than the actual observed mobilities .
We are forced to the conclusion that the charge cannot have The results for the positive mobilities are exhibited in the curves shown in fig. 8 ; the curve for pure hydrogen is based on Zeleny 's values , and several points on the for pure methyl iodide have been calculated on the assumption of the validity of the law .
constant .
In order to interpret the results , let us consider bhe particular curve corresponding to 7 cm .
of methyl iodide .
It is evident that the addition of a small quantity of hydrogen does not appreciably alter the mobility ; had the vapour ions been unstable and transferred their charge to the hydrogen molecules we would have expected a marked increase in mobility due to the small mass of the new carrier .
When further hydrogen is added the mobility decreases only very slightly ; in fact , although the pressure changes from 7 to 76 cm .
the mobility only decreases from cm . .
to 2 From bhis result we deduce that as hydrogen is added to the methyl iodide the charge tends more and more to be carried as a ion , although the methyl iodide molecule can accompany the charge to quite an appreciable extent even in the presence of a considerable quantity of hydrogen .
From an inspection of the curves it can also be seen that the mobilities approximate more quickly to those of pure hydrogen the smaller the initial quantity of methyl iodide .
Considerations of a similar kind apply to the negative ions .
1909 .
] Electricity through Gaseous ixtures .
The preceding considerations with to the transference of charge from the vapour molecule to the hydrogen molecule depend ultimately on the applicabihty of Maxwell 's formula for the mean free path of a single uncharged molecule of mass in a medium consisting of molecules each of mass denoting the number of molecules per unit volume , and the sum of the radii of the two kinds of moleoules .
The application of this expression to the case of the diffusion of gases has been made by O. Meyer , calculated the diffusion coefficients for a series of gases by a method involving this formula , and obtained a striking agreement with observation .
If denote the mean free path of the methyl iodide molecule through hydrogen at any pressure , and the mean free path of the methyl iodide molecule through methyl iodide at this pressure , and if we regard the above formula as valid we deduce where and denote the radii of the vapour and gas molecule respectively .
Taking as before , we obtain approximately .
Now , the mobility at constant temperature is proportional to the mean free path , provided the ion is unaltered in mass and dimensions , as is readily seen from Langevin 's expression for the mobility : , where denotes the mass of the ion , its mean free path through the gas , and its mean velocity of thermal agitation .
It follows , therefore , that if the ion formed in pure methyl iodide at any pressure be unaltered increased in mass and dimensions by the addition of any quantity of hydrogen , the curves for the mobilities shown in fig. 8 would all lie nearer the axes than the mobility curve for pure methyl iodide .
The actual positions of these curves as determined experimentally imply that the ion diminishes in mass and dimensions as the hydrogen is added .
If the ion is then to be regarded as consisting of a cluster of molecules , it is necessary to postulate the original transference of charge by either of the methods indicated above and , in addition , the decrease of the ion in mass and dimensions as the hydrogen is added ; this might possibly arise if we regard the addition of hydrogen as decreasing the time during which the ionic cluster involves a methyl iodide molecule , so that when a considerable quantity of hydrogen is present the ion would be approximately a cluster of hydrogen molecules .
Another explanation as to the nature of the passage of the electricity through the medium is given in the next paragraph ; this view is capable of explaining in a simple manner all the preceding results and , moreover , is in harmony with other * Kinetic Theory of Gases , ' 2nd Edition ( English ) , p. 274 .
VOL. LXXXII.\mdash ; A. 1909 .
] Electricity through Mixtures .
the molecules of certain vapours ( e.g. , water , alcohol , acetone ) tend to hold the charge in association with them much more strongly than the molecules of the heavier vapours such as methyl iodide .
Thus if a small quantity of water-vapour be mixed with carbon dioxide at one atmosphere the charge ( both positive and negative ) would be carried for a considerable portion of the path by the water molecules ; owing to the relatively small mass of the molecule of water we would expect an increase in the mobility of the positive ion , but in the case of the negative ion this increase might be more than counteracted by the shortening of the period during which the corpuscle exists in the free state , so that on the whole the mobility of the negative ion would be decreased .
Similar reasoning would explain , although merely descriptively , all the observed phenomena with regard to the effect of small quantities of vapours on the mobilities of gaseous ions .
It is of interest to record that Prof. Sir J. J. Thomson*has recently advanced the theory that the act of ionisation consists in the expulsion from the molecule of both a positive and a negative unit of electricity ; such a view is quite in accordance with all the experimental results described in this paper .
I take this opportunity of expressing my gratitude to Sir J. J. Thomson for the inspiring interest he has manifested throughout the research and for many valuable gestions .
* Phil. Mag vol. 16 , p. 685 , 1908 .

