KXI'KKl.MKNTS WITH lONIZKIl All;. 



6. Ftirther data. — Moreover, the data of this table ai'e uniformly smaller, 

 scarcely I- or \ of the values of table 1, due possibly to the lower temperature of the 

 present case. They are insufficiently definite to suggest the form of the locus. 

 New observations wei-e therefore needed and were made in the same way, with the 

 results given in table 3. Tiie condenser was slightly modified, consisting in this 

 case of a sphere of wire gauze about 1 centim. in diameter, containing a much 



