982 
EEV. T. E. EOBINSON ON SPECTEA OE ELECTEIC LIGHT. 
None of these, however, are peculiar, except that Nos. 11, 17, and 54* appear as 
double. 
That the presence of a new body in the circuit developes oftener than originates lines, 
is further shown by moistening the electrodes with a solution of some salt ; for instance, 
chloride of barium f. In trying it, platinum electrodes (carefully washed uith nitric 
acid and distilled water) were lapped with clean cotton thread to retain the fluid, which, 
however, soon evaporates by the heat of the discharge. The change of the spectrum is 
almost startling from the instantaneous and intense development of many lines, and the 
increased lustre of some which before had been noted as ^^s. The most conspicuous 
<11 O ^ 
a. Very bright 32 38T 
a'. * very intense 40 '0 
j8'. -X- dazzling'' 49T 
7. Very bright 54’ 7 
No. 16. Orange intense . . . 56 ‘3 
No. 17. Very bright .... 57’7 
No. 34. * very intense .... 33 18'4 
Nothing striking till then 
No. 47. * brighter than ^ . . 37’4 
No. 48. Almost * 39-4 
A. of exceeding intensity . . 34 2'3 
X'. Very bright 4' 6 
No. 80. Extremely bright ... 40 ‘6 
All these are found in the platinum air spectrum, though in most cases mth a totally 
different aspect ; so that the barium makes no change in the place of the lines, but only 
in their brightness. 
2. That the peculiar character of a line is modified by other circumstances than the 
chemical nature of the bodies engaged in producing it, appears from the difference 
between the spectrum at common-pressure, the transition spectrum, and that in rarefied 
gases, to which I have already referred. The chemical conditions are unchanged ; but 
at first sight nothing is more dissimilar than the three; the first with its numerous 
bright ifes; the second a shadow rather than a spectrum, with a bare suspicion of a few 
lines ; and the R. one much brighter, but with a seeming discrepancy of lines that marks 
it as peculiar. The R. has, on an average, 0'53 of the C.P. lines; it extends as far in 
the wolet, but is cut off in the red part. There, especially in N, are often found a set 
of equal bands reaching from No. 11 to No. 19 inclusive ; there is also another set, that 
* This line is always double with S^C prisms, 
t It is desirable to ascertain whether this process gives a spectrum always identical with that given by 
electrodes of the metal contained in the salt. The difference of aggregation and the presence of water may 
modify it considerably. 
