356 G. Hinrichs on Spectral Lines. 
The difference between the first two lines is too small to war- 
rant any conclusion drawn from the same. 
Group IV. d,=1°46. In the yellow. 
I Ww D i wd E 
8d =—-:1235°0 = 555820 57 g_~SC«#282-00 00 
4e = -:1229°6 = 55 87°7 17. gy. OUST SE eG 
9d 12283 55894 4g g . 958030 +10 
5d 1224°7 5593°6 3-9 g 559368 —'08 
5d 1221°6 5597°5 0-5 g 5596-60 —'10 
3c =—s-«1219°2 = 56000 aa 1 559952 "42 
Sd = 12178 += 5601-0 560098 +02 
These intervals form the following simple series: 
2 
3 
Nisetreaentynnennioned 
or 4 4 2 2 ae 
ee 
or 4 + 4 1 
Group V. d,=6°57. In the orange. 
I K Ww D t Ww’ E 
2e 8949  6102°0 19-0 3 6102°0 00 
4b Soro C1210 Gg aekt 7 
56 863°9 6161-0 6-7 1 6161-13 —13 
3d 860°2 6167-7 6167°70 00 
Grovr VI, P 
There are too few observations for a very accurate determina- q 
tion of the corresponding wave-lengths of this group in the red 
part of the spectrum. The difference of 2°3 K as found in our 
preliminary investigation corresponds to 623 wave-length in 
this part of the spectrum. : 
Comparing the different groups we need not expect the differ- 
ence d itself to be the same for all groups; but such a multi- 
lum thereof as the intervals indicate, ought to be the same. 
hus we find 
Group I, d,=3°3 2d ,—=6°60 
« TI, d,=116 6d = 6-96 
ey AV, d xel46 4d,—5°84 
ee Vy O57 1d,=6°57 
“ Vi, d,=6-23 ge 6'23 
ld 
the mean of which is =—6°44 
As the above numbers go, we see that the interval ap- 
roaches equality, but is not strictly so. Yet it may well be 
rne in mind that the greatest deviation either above or below 
the given mean value amounts to only half a unit; so tha 7 
though the equality of the interval cannot be considered as de- 
monstrated, yet the inequality is not demonstrated either. __ 
It is now necessary to consider the single lines given by Kireh- 
