14 



MR. O. W. RICHARDSON ON THE IONISATION 



B 



changes must have affected to some extent the results shown in fig. 3. The practically 

 exact coincidence of the two curves might merely imply that the observations were 

 taken at about the same rate in the two cases, so that the errors were about the same 

 in each case. A check on this was, of course, afforded by the fact that values of the 

 leak were only retained after they ceased to vary with the time ; but partly owing to 

 the great length of time required for equilibrium to be established, and partly owing 

 to the invariable presence of irregularities of another kind, which will be considered 

 later, it was very difficult to be sure that equilibrium had been attained in any 

 specified case. For this reason it was thought desirable to have some further means 

 of checking the extent to which these influences affected the measurements. 



This was done by gradually increasing the oxygen pressure instead of diminishing 

 it, as happened by taking readings after successive strokes of the pump. It is 

 evident that under these conditions the pressure time lag will have the opposite 

 effect to what it had when the pressure was being reduced, so that the lack of 

 coincidence between the curves obtained with increasing and diminishing pressure 



will give a measure of the extent to which the 

 lag effect has not been eliminated. What was 

 required, then, was an arrangement which would 

 deliver small adjustable quantities of oxygen 

 into the main apparatus. 



To do this, the apparatus shown in fig. 4 was 

 devised. The principle of the method is to allow 

 gas to flow into the apparatus through a very 

 long narrow capillary tube under an adjustable 

 difference of pressure for varying times. In the 

 figure this part of the apparatus is shown 

 together with the arrangement for furnishing 

 the oxygen. The bulbous tube A contained 

 pure dry potassium permanganate ; by heating a 

 small portion of it for variable lengths of time, 

 any desired quantity of pure oxygen could be 

 set free. Such a tube may be used time after 

 time for long periods. B is a glass wool plug to 

 stop the manganese dioxide dust which is pro- 

 duced when A is heated. C is a drawn-out glass 

 tube which can be broken to let down the 

 vacuum in this part of the apparatus if required. It is convenient to have the 

 permanganate tube so that it can be shut off from the rest of the apparatus. This 

 may be done by means of the mercury cut off I). At F is the fine capillary tube, 

 bent as shown to economise space, through which the gas flowed into the main 

 apparatus at H. This tube was about 140 centims. long, and was the narrowest 



Fig. 4. 



