56 



Proceedings of the Royal Irish Academy. 

 (c) Time— 182 sees. 



Mea 



— 



— 





00035 



•0023 



•0016 



•00071 



•00037 



•0023 



•0015 



•00065 



•0003-1 



— 



— 



— 



•00031 



0023 



•0015 



•00068 



•00035 



(</) Time— 409 sees. 





— 



— 



•00031 







•0012 



•00018 



•00030 







0010 



■00052 



•00026 







0012 



•00(117 



■00028 





Means. ... 



■0011 



II!' 



•00029 





(<•) Time— 900 sees. 





— 



— 







00015 





0016 



■00068 



00031 



•00011 





— 



— 



■00031 



•00011 





— 



•00055 



00032 



00015 





— 



— 



— 



■00011 



.... 



•0015 



•00062 



00031 



•( 1 1 



From the numbers given in Table III we gel further confirmation of the 

 system of division of the ions into classes. Each class has a definite mobility, 

 and in tl - ol experiments with.widdy difl'cirnt conditions, no ion is 



observed which cannot be classified. 



It will be noticed in Table III thai foi the time M)9 seconds the mobility 

 values are abnormally low. This was probably because a very Blow air stream 

 was used. Some additional experiments were made to test whether these 

 low values indicated a true variation of mobility with hup-. In the above 



aliments with a time-interval of 409 seconds, the velocity of the air stream 

 was 7*8 cms. pel second. Air velocities of 18*2 cms. and 1 t'6 cms. per second 

 were now used and a volume of tubing which gave time-intervals of 433 

 seconds and 540 Beconds respectively. Ei^ht good determinations of the last 

 ion of Table III (d) gave a mean value of -OOO'T. The low values for time 

 409 seconds in the table are therefore probably due to the slow air stream, 



