]\gt ; in comparison with the velocity under an electric force .
It was suggested ( loc. cit. , p. 469 ) that this could be explained on the supposition that the ions were not in " " thermal equilibrium\ldquo ; with the molecules of the gas , but that their mean kinetic energy exceeded that of an equal number of gas ] ecules in the ratio to unity , where depends on the pressure and the electric force .
Townsend mentioned a possible means by which this abnormal energy might ] arise , namely , that ' extra velocity acquired by an ion ih an interval * J. S. Townsend , ' Boy .
Soc. Proc 1908 , , p. 464 .
, energy might accumulate for some time .
The effect has been further studied by Haselfoot , *both observers showing that is , at any rate approxi ; mately , a function of only , where X is the electric force and the pressure of the gas .
It seemed to the author that it would be desirable to investigate the matter quantitatively on the basis of the kinetic theory of gases , since we are here dealing with the validity or non-validity of the law of equipartition of energy .
It is shown in this paper that it is easy enough to obtain theoretical support for the assumption of abnormal kinetic energy ; in fact , the values of predicted by pure theory are considerably in excess of those actually observed .
In order to prevent misunderstanding , it may be stated that the law of equipartition of in the kinetic theory is only proved to be true when the molecules are left to themselves in the absence of external 1 forces .
2 .
The most satisfactory way of treating these questions is that developed by Maxwell in his later papers , and carried out by him on the assumption of an inverse fifth-power law of repulsion between molecules .
Where not otherwise stated we shall follow the notation used by Jeans in his account of Maxwell 's theory .
Let mber of ions per cubic centimetre of the gas , small in comparison with Number of molecules of gas per cubic centi1netre .
Electric force .
Absolute ture of the Velocity of an ion .
the mass-velocity of a group of ions .
Writing , w , is called the velocity of agitation of an ion , so that Velocity of agitation of a gas molecule .
Mass .
ion and molecule respectively .
Average kinetic energy of agitation of a gas molecule .
average kinetic energy of agitation of the ions in a specified small region .
C. E. Haselfoot , ' Roy .
Soc. Proc 1912 , , vol. 87 , p. 350 .
J. C. Maxwell , ' PhiL Trans , ' vol. 167 ; 'Collected Papers , ' , p. 26 .
'Dynamical Theory of Gases , ' Chap. XV .
3 .
The immediate problem with which we are concerned is to find the kinetic energy in a stream of ions advancing in a fixed direction , for example } the axis of , the motion depending on the co-ordinate alone .
Then constant , ( 9 ) , ( 10 ) , ( 11 ) , ( 12 ) .
( 13 ) * Jeans , p. 277 .
'Collected Papers , ' vol. 2 , pp. 47-48 .
motion tends wtream h. some distance .
In this case Substituting for X from this equation in ( 11 ) , ( 12 ) , and ( 13 ) , we find S These latter equations yield on addition ; ( 16 ) or , if is the " " root-mean-square value\ldquo ; of the velocity of agitation of the gas molecules , 4 .
The inverse fifth power law has the advantage of allowing a perfectly rigorous solution of the present problem , free from all considerations of approximation .
As , however , the formula ( 17 ) is not in agreement with experiment it is desirable to consider another law of force , namely , that both ion and molecule are elastic .
Some useful information is given by equations ( 15 ) as to the distribution of velocities of agitation in val.ious directions .
In general are unequal , and their differences are of the order of magnitude of .
In two cases , however , the differences are small in comparison with the absolute value of any of the quantities ; namely , when is nearly unity , and when is small in comparison with M. We shall , therefore , not be far wrong in Maxwell 's law of distribution as a first approximation .
Let represent the fraction of the whole number of ions which have velocities of agitation in the range and : then Maxwell 's law is expressed by .
( 18 ) There is no doubt as to the corresponding function for the molecules of the gas , namely , We shall suppose for the sake of generality that the collision of ion and molecule is not perfectly elastic , an assumption which allows roughly for a possible loss of energy on collision , it renders the validity of ( 18 ) more doubtful than it might otherwise be .
Under these conditions

