110 Proceedings of the Royal Irish Academy. 



The collision frequency between small ions and uncharged nuclei of 

 radius a, present in concentrations N and «,„ is ira-Nn^v., where u is the 

 velocity of agitation of the small ion. The coefficient of combination is 

 therefore ira-u. If for the coefficient the value 5 x 10"" is taken, we have for 

 the radius of the nucleus 



X 10-" j^.Qo ^ -^Q.5 ^^^^_ 



3-14 X 1-5 X 10* 



Thus again we find that the experimental results force us to regard the 

 nuclei and large ions as, if anything, greater than 1 x IQ-' cms. in radius. 

 From the point of view of the loose grouping theory of the large ion, this 

 itself does not offer any difficult}'. In fact, it is to be expected that the 

 volume occupied by the loose group should be large in comparison with its 

 mass. There remains the difficulty of the recombination rate of the nuclei 

 among themselves. If this value for the radius of the nucleus be accepted, 

 we have to assume, as onr previous calculations show, that only a small 

 fraction of the collisions between nuclei result in coalescence. There is also 

 the difficulty, previously mentioned, as to the disposal of tlie products of 

 recombination. 



The Charge on the Large Ion. 



All arguments as to the size of the large ion founded on mobility results 

 are complicated by the question of multiple charges on the ions. Charges 

 vary from one to six electrons under different conditions, but no correspond- 

 ing variations in mobility are easily made apparent. What we are accustomed 

 to regard as the stairdard mobility ("00033) is nearly always found, and there 

 is often associated with it an ion of mobility about three times tliis value. 

 Such other variations from the standard mobility as are found, for example, 

 in the experiments with SOj, NHs, and Na CI, or in the heating experiments, 

 cannot be associated with variation in cliarge. 



As the apparatus which we have used for the recombination observations 

 is very suitable for the determination of the charge on the ions, we have 

 carried out some experiments on this point. The gas containing the ions is 

 passed through the apparatus at a uniform rate until everything has become 

 steady. The total number of nuclei per e.c. is then found by means of the 

 Aitken apparatus. A field of 800 volts is applied to A so that all the ions 

 are removed. Tlie Aitken apparatus then gives the number of uncharged 

 nuclei per e.c. If the electrometer is connected to the central electrode at 

 A, the total numljer of electronic charges of one sign per e.c. of gas can be 



