1317 
This is still more evident from fig. 1, where for each group we have 
75 
70 | if 
a i 
60 | 
: An a Drees Group I 
DARMEN ee ee) eee Group II 
8 | ——— Group IV 
: | == Group V 
Ci 
o 
Number of lines in 9%Jp 
8 & 
bo 
o 
15 
10 | ui 
5 aes 
| aS 
N 2 
S 
clearly indicated the percentage of the lines of each group that 
shows a Z.E. (expressed in 0.001 AU) from 0—50, 50—100, etc. 
From this we see that group I has a sharp maximum between 350 
and 400 (73°/, of all lines). Further that also group II shows a 
maximum in that same region though much less sharp (27° ,) and 
that this group shows already distinctly some inclination for greater 
Zrrmar-effects. The groups IV and V have each two equivalent 
maxima viz. for group TV between 450 and 500 and between 700 
and 750 each of 21°/, with a lower maximum (14°/,) between them 
and for group V between 500 and 550 and between 650 and 700 
both of 16°/, while between these lie still 24°/, of all the lines of 
this group. 
Considering now the observations on the pressure-effect, then 
we directly remark that for the lines of one and the same group 
the magnitude of this effect is very different. That for one and the 
same line observations of the are or of the spark-spectrum would 
give somewhat different values was to be expected a priori, but 
that observations on the same line in the same spectrum give such 
different values of the pressure-effect per atmosphere seems very 
strange. 
From the observations of Binnam at 5 atmospheres over-pressure 
(B) we find a higher value for the pressure-effect than from those 
of the same author at 10 atmospheres over-pressure (B). We may 
therefore directly conelude : 
