1321 
From the figure is seen that the ratios of group IV are distributed 
over a wide region without showing anywhere a trace of a 
maximum. 
Group V shows a very small region in which all values of the 
ratios are lying, viz. between 27 and 52. Except for the first line 
for which no Z.E. has been determined, the values of B,, D, and 
D, show here a very great and rather good agreement among 
each other. If therefore in any group there can be question of a 
direct proportionality between Z.E. and Pr. E, it will be for the 
lines of group V. Unfortunately Bituam has not extended his 
quantitative observations over more lines of fhis group; else we 
might perhaps have been able to verify this conclusion still better. 
After this explication of the figure we may use still further the 
VN ZL 
numbers given as “Means.” Assembling the means of — ,—,—, = 
B Beep Dy: 
LE Z 
and ——| viz. the mean of = sl in a table, and reducing those 
Fy Bor De 5 
of the first group all to one, we obtain: 
a Z 7 Z g ZE 
P B, B, D, D> DE 
= eas ze er 
I 1 1 1 1 1 
ll 0.84 0.93 0.85 0.82 | 0.90 
ll 0.34 0.42 0.34 | 0.36 
Iv 0.22 0.40 0.68 0.93 |- 0.48 
v 0.19 0.24 0.14 0.18 
From this table it is evident that in the study of the ratio of 
ZwrMAN-effect and pressure-effect the distinction of the different groups 
gives as pronounced results as in the investigation of the pressure- 
effect alone. For the groups | and II this ratio is greatest, while 
group V is characterized by a very small ratio. Though the groups 
I and II show great agreement, so that Binwam') feels inclined to 
combine them to one single group, we see here that in the com- 
parison of the two effects the difference is still considerable. As for 
these two groups a sufficient number of observations was at our 
disposal, we can look upon this difference as being a “real” one, 
1) 1. c. p. 368, 
