348 
nevertheless tabulate the values of the product vA for all gases 
we have examined; this procedure after all is extremely serviceable 
in helping us to understand the correlation of the values of A for 
different substances. 
Before stating the result some preliminary remarks may be ne¬ 
cessary. It must be remembered that all values of the product vA 
are given (throughout) in terms of the unit: 10“ 7 8 cm 2 . The next 
remark called for is that the number v is not wholly determinate 
in every case; it is a well-known fact (on which it is needless to 
dwell here) that the number v has never been defined satisfactorily 
and may perhaps be undefinable, precisely in the case of some of 
the simplest gaseous molecules known in Chemistry. Thus the value 
of v adopted for CO, S0 2 , N 2 0, at least to a certain extent, must 
be admitted to be arbitrary. 
In the case of air the constant A is given by 
(3) - = «i -j- a 2 
(see § 18.); and a 2 are given by (2), § 14. Let us suppose M 1 
molecules of oxygen and M 2 molecules of nitrogen are contained 
in unit volume of air at standard temperature and pressure; and let 
(4) M 1 = \ (M 1 + M 2 ); M 2 = h 2 (i¥i + >/ 2 ). 
If we put: 
(5) ! = a ! M 1 ; N 2 = a 2 M 2 
and if for oxygen and nitrogen electrons we assume 
(6) 2 an d e i — e 2 
we find from (3) and (4) 
(7) i = + 
thus atmospheric air will behave as a single medium and its a will be 
(8) a = % Äj -f- a 2 h 2 . 
The legitimacy of this conclusion rests on the assumptions explai¬ 
ned in § 18. Taking 
= 4; v 2 = 6; /^ = 0,209; h 2 == 0,791 
and substituting v 1 for a 1? v 2 for a 2 , we get v — 5,582. 
