354 G. Hinrichs on Spectral Lines, 
§ 15. The preceding tables may be condensed to the following: 
Substance, d Values of 7. 
Hydrogen, 169 10 3 
Chlorine, 47 5 9 
Bromine, 25 15 1 3 
Iodine, 97 8 ee ee ee 
: on 17 84 56 
or 11 12 8 
or 243 3 2 
Nitrogen, 65 8 
Oxyen, 1373 6 1 6 
Mercury, 25 x 12 14 
Terchlorid of phosphorus, 36 13 12 
Bichlorid of silicon, RRS 8 
Bichlorid of tin, 10 65 21 20° 
Carbonic acid, 10 41 69 12 
$16. It can hardly be assumed that the determinations of 
Pliicker are as accurate as those of Ditscheiner; the latter are, 
as will be shown, not always reliable in the last figure. Hence 
the agreement between observation and our calculation 1s in 
most cases astonishingly close; thus for hydrogen, where d= 
169 there is only one single deviation, and this amounts to only 
three units; the chlorine spectrum shows as good as no devia- 
tion at all, only one unit for d=47 on a range of 700. It is similar 
for bromine, oxygen, nitrogen, and mercury. Only iodine devi- 
ates for one single line (9) more than is proper, especially as d 
is only 8; but then this being the only instance, it may be due 
to a greater error in the observation of 6. 
It may perhaps be objected, that only few lines are taken for 
each element. But we have taken ail lines for which determin- 
ations could be found, and though they may be few in number, 
they are ranging through the greater part of the spectrum. — 
uming that the other metalloids will correspond in this as 
they do in all other respects, we may express the above by the 
following law: the wave-lengths of the dark lines in the spectra of 
the several metalloids differ from one another by simple multiple of @ 
certain number, d, peculiar for each element, and the spectra of the . 
binary compounds given do not show so simple multiples, but 
the differences are yet expressible in multiples of a certain num- 
ber for each compound. 
_ The above law is substantially the same as the laws deduced 
in the preliminary investigation. , 
17. In all the foregoing tables we have made use of Pliickers 
observations, which are only known to us in their final result, 
The following values of W will be taken from Ditscheiner, and 
as his memoir js at hand, we may first ascertain the degree of 
accuracy of his observations. 
