341 
the mean 
is A = 3,90. Thus ScheeFs results 
seem to corroborate 
what has been deduced from Mascart’s data. 
From ScheeFs a 
and b values we find: 2 0 
= 0,88. 
§ 20. 
Carbon 
monoxide. For CO, Perreau’s and MascarFs 
data are 
as follows: 
Perreau 
Mascart 
a) 
2 == 4,677 
... v= 1,0003388 . . 
, . — 
b) 
4,800 
3382 . 
. 1,0003384 
o) 
5,085 
3368 . 
3370 
d) 
5,378 
3357 . 
3358 
e) 
5,896 
3342 . 
3345 
/) 
6,438 
3328 . 
3334 
From (a) and (f) (Perreau) . . . 
A == 3.67 
... (b) and (e) (Perreau) . . . 
3,67 
... (c) and (d) (Perreau) . . . 
3,68 
From ( b ) and (e) (Mascart) . . . 
3,53 
. . . (c) and {d) (Mascart) . . . 
3,88 
Takin 
g 4 = 3,67 
we find from Perreau's 
numbers D = 4594; 
this gives 
K—- 1,000653. The observed values 
are higher than this: 
K-- 
= 1,000690 (Boltzmann); = 1,000695 (Klemencic). 
§ 21. 
Carbon 
dioxide. In the following table the measure- 
ments due to Perreau and to Mascart are collected. The results, 
though discordant, lead to approximately equal estimates of the dis- 
persion effect. 
Perreau 
Mascart 
a) 
2 = 4,677 
... v= 1,0004550 . 
. . — 
b) 
4,800 
. . . 4544 . 
. . 1,0004581 
c ) 
5,085 
4530 . 
4557 
d) 
5,378 
4518 . 
4550 
e) 
5,896 
4502 . 
4538 
/) 
6,438 
4487 . 
4526 
From (a) 
and (f) (Perreau) . . . 
A = 2,15 
...(b) 
and (e)' (Perreau) . . . 
2,12 
... (c) 
and (d) (Perreau) . . . 
2,17 
From (b) 
and (e) (Mascart) . . . 
2,12 
... {b) 
and (f) (Mascart) . . . 
' 2,06 
