4 Proceedings of the Royal Irish A cademy. 



•00015, which could not be found a few seconds after the formation of the 

 ions. The same question which we discussed in the case of the alcohol ions 

 arises. We may account for the appearance of the ion of mobility -00015 by 

 supposing that it was present at the very beginning as a small fraction of the 

 total ionisation, and that the faster decay of the smaller ions brought it into 

 prominence. Or, we may suppose that it was not present at first, but that it 

 was formed from some grouping of the smaller ions. It is dithcult to decide 

 which is the correct view. As was pointed out in the previous paper, it is 

 difficult, with this method of finding mobilities, to measure and compare the 

 percentages of the different classes present under different conditions. The 

 evidence in this case would not prevent us from thinking that the ion of 

 mobility 'OOOlo is present originally. In the case of the alcohol ions, we 

 thought that a more likely explanation of the appearance of this ion was 

 that it was formed as time went on. Perhaps the most satisfactory way of 

 looking at this question is to suppose that in both cases the large ions are 

 formed from the small ions, and that the rate of formation depends on the 

 source of the ionisation and on otiier conditions. If the grouping of the more 

 mobile ions to form tlie large ions took place very rapidly, we would have a 

 system of ionisation very similar to that contemplated in our first theory. 

 The evidence on the whole indicates that tlie large ions are formed by group- 

 ing, and tiiat the grouping may proceed at widely different rates in different 

 cases. In the case of phosphorus, the grouping takes place so (juickly that 

 we cannot assert that the large ion of mobility -00015 is not present in small 

 quantity after a few seconds. On this view there is no essential difference 

 between the ionisation due to phosphorus and that due to alcoliol, although 

 in the former case we can detect the ion of mobility -00031 after 1-3 seconds, 

 whilst in the latter case this ion was not observed until over a minute had 

 elapsed after bubbling. 



In the experiments with time- intervals 1-3 sees., 4-6 sees., and 16 sees., 

 the ion of mobility 00063 was very prominent. It was present in greater 

 quantity than any other ion, and gave about one-third of the ionisation. In 

 some of the observations the quantity of the ions of mobilities -0024 and -0041 

 was very small. At times, indeed, it was difficult to be certain of their 

 presence. 



A mobility tube havmg a short terminal was used so that the more mobile 

 ions could be more conveniently examined. The time-interval was reduced 

 to -8 second with the same object. The glass tube which contained the 

 phosphorus was covered with tuifoil and placed in metallic connexion with 

 the mobility tube. The object of this arrangement was to guard against the 

 possibility of the smaller ions being turned back by the field which they 



