712 Prof. H. A. Wilson on the Velocity of the 



one monovalent positive ion, the result just mentioned makes 

 it probable that the same thing happens in the vapours. 



Prof. 0. W. Richardson* has recently measured the ratio 

 of the charge e to the mass in for the positive ions of vapours 

 of the sulphates of all the alkali metals, and finds it equal to 

 the value which obtains in solutions. This makes it yqvj 

 probable that the positive ions are metal atoms. The fact 

 that the haloid salt vapours give ions having the same 

 velocities as the oxysalts in flames, shows that ail salts of any 

 one meial give ions identical in nature. 



The equality of the velocities of a lithium ion and a caesium 

 ion is difficult to explain on the view that they are 

 simply single atoms, for we should expect the velocity to 

 depend on the atomic weight. The main object of this paper 

 is to point a way out of this difficulty. 



In my experiments two electrodes were placed one above 

 the other in a Bunsen flame, and the current between them 

 was measured. If the upper electrode was positively charged 

 and a bead of salt was placed just below it, it was found that 

 the current was not appreciably increased by the salt unless 

 the potential difference between the electrodes was greater 

 than about 100 volts. This was taken to mean that 100 volts 

 was just enough to make the positive ions move down the 

 flame. The potential gradient in the flame is nearly uniform 

 except near the electrodes, so that the current density (z) is 

 given by i = en(y 1 -\-v^), where n is the number of ions of 

 either sign per c.c, v 1 and r 2 the velocities of the positive 

 and negative ions. If the gas is moving upwards with 

 velocity u and X denotes the electric intensity, then we 

 have 



Vi = &iX — u 



so that the current is equal to enX.(k 1 4- £ 2 ), and is independent 

 of u. 



When the upper part of the flame is filled with salt vapour 

 n will be much larger in that part than in the re^t of the 

 flame, so that for a given current X will be proportionally 

 smaller. This diminution of X, however, does not lead to 

 an appreciable increase in the current, when the upper elec- 

 trode is positive, because nearly all the resistance to the 

 passage of the current is close to the negative electrode, 

 where the greater part of the fall of potential takes place. 



* Phil. Mag. Dec. 1910. 



