EEY. T. E. EOBINSON ON SPECTEA OE ELECTEIC LIGHT. 
979 
the eye be directed to the negative boundary of the spectrum, two brilliant points arc 
seen there not found in the ordinary H band. At the same time they may be due to 
the grosser metallic vapour which abounds there ; for Te volatilizes under the discharge 
more than any other metal which I have tried. Iron in N, if the pressure be diminished 
a few inches, shows a similar band, though the other characters of the C.P. spectrum 
and discharge are unchanged. Zinc in CO, C.P., has one of the same sort 3' broad. 
Nos. 51 and 52 are a well-marked pair ; but a much more remarkable group is that 
from No. 54 ; to No. 59 yJ’. Of these, 55 and 58 are very narrow, and, though bright, 
are often overpowered by the splendour of their neighbours ; the first has a still fainter 
follower seen in the S^C prisms, which also show each of the brilliant ones to be com- 
pound, t and fi double, yJ' (which begins the ■violet) to consist of six fine bright lines. In 
thu'ty-two cases y and were observed as one, or as close double. When taken sepa- 
rately, their distance was found 0'‘96 ; and hence the place of y, should be increased, and 
that of y! diminished by 0'‘14. H is unfavourable to this group, which is most splendid 
in O and CO, except in Cu, CO, C.P., when it is but a shadow of itself. 
The ground is covered with very fine lines from No. 61 to No. 68 in the CO, C.P., 
spectra of Cu, Pd, Fe, Bi, Sb, and Mg. 
No. 72 j', besides being important in the spectra of other gases, is apparently developed 
in H, C.P., to the broad violet band, the third of the three brilliant ones of that gas; 
22 of H, C.P., give its place 34° 19'-82. It is not nearly so luminous as No. 49 is in H, 
but is quite as broad and even more undefined. Unlike other violet bands, this is the 
first to disappear on rarefying the gas. I have occasionally seen fine lines in it, which, 
however, do not seem to belong to it. 
No. 73 is in the place of G, and No. 85 very nearly in that of H. 
If the individual spectra be compared with this Table, the result will, I think, throw 
some light on the questions which I mentioned at the beginning of this paper. 
1. As to the essential connexion of each line of a spectrum with the chemical nature 
of some one metal, or some one gaseous medium, I think these observations are against 
it. For instance, if any line, say No. 22 be found with electrodes of aluminium in all 
the gases which we have examined, our first inference might be that it is an aluminium 
line. But wiien we find the same thing is true of nickel, palladium, antimony, and 
magnesium, and that all the other metals have it in some of their spectra, we must con- 
clude that it does not belong exclusively to any metal ; nor can it be considered a gas 
line. In N it is indeed found with all the twenty-two electrodes ; but it also occurs nine 
times in O, fourteen in H, as well as in the vapours of mercury, phosphorus, and bisulphuret 
of carbon. There are many similar cases ; and though in general the lines of this Table 
are not seen with all the metals, yet it will be found that, with respect to gases (omitting 
No. I, which on account of its bright neighbours is seldom seen), out of the whole eighty- 
five only Nos. 5, 6, 9, 73, 79, and 85 are wanting* in H, C.P. ; only No. 6 in CO, and 
that of these all but No. 5 are found in the E. spectra, and none are wanting in A, N, 
* That is, are not visible with my means of observing. 
6 s 
WDCCCLXII. 
