EEV. T. E. EOBINSON ON SPECTEA OE ELECTEIC LIGHT. 
951 
except H, C.P. (which was taken with Duboscq, and at the negative boundary) were 
with Merz, but with the improved telescope. As first observed, and at the centre, this 
last showed only eight lines. 
It will be remarked that all the lines in the C.P. of nitrogen, and all but one in that 
of oxygen, are found in air, as might be expected ; often a line is common to the three, 
and then that in air is of intermediate character. But the same line is also often found 
in the other two gases. It is generally supposed that this indicates the metallic origin 
of that line ; but it will be found that many occur, not only in all these gases, but with 
all or nearly all the electrodes which I have tried. I shall return to this at another 
time, now making a few remarks on those before us. 
The pair a, a! are of almost universal occurrence at the origin of the spectrum ; a has 
close before it a narrow but bright companion, whose place has sometimes been taken ; 
but outside of that there are merely shadowy traces scarcely ever bright enough to be 
bisected. In a very few cases, however, one of them becomes predominant, e. g. silver 
and iron. In H (hydrogen) spectra, one of this pair acquires an intense brilliancy, 
which is peculiar to this gas, while the other fades away. It is to be noted that a is 
exactly in the place of the solar line C. These and /3, /3' are in the red. No. 7, y is an 
orange band, nearly but, I think, not exactly in the place of D. When viewed with the 
combination 2BS.C, it is double in all cases which I have examined ; but the distance of 
the centres of its components is fully twice that of the components of D. No. 11, h is 
yellow, generally extremely brilliant, except in one ; it is seen in 2BS.C double, the 
second one being about half the breadth of the preceding. Before it is a narrow orange 
fine, and before that No. 10, which seems to attend it constantly. Nos. 14 s, the origin 
of green, 15 g' and 16 g" are of constant recurrence. With the higher prism-power the 
first and third are each double, and some between g' and g", of which one (as here in 
air, C.P.) has been occasionally observed. No. 18, was at first observed as a bright 
cloudy band ; 2BS.C shows that the whole of that region is covered with close lines, of 
which one is more conspicuous than the rest. In CO many more are conspicuous. 
No. 21, Ti is very brilliant (except in oxygen and CO, when its following companion is 
the brightest). With 2B8.C it is close double of two equal*. This is also the case 
with No. 24, at the boundary between green and blue, which is, as a rule, the most 
intensely bright in all the C.P. spectra, in many of which its light is almost blinding. 
In E. it is always faint though present. It is preceded by two narrow ones of unequal 
brightness, which seem to form with it a system. Another system seems to be found 
from 32 to 36, remarkable for its beauty and peculiar character; is vivid blue, close 
double of equals, broad and cloudy, begins the violet. With 2BS.C x, is composed of 
two further apart, and x!' of six. It is less developed in nitrogen and H ; 'k 37’40 was 
at first observed as a band; in CO the whole of its vicinity is covered with narrow 
violet lines. The pair 28, 29, and the triplet /a, |M/', are also of very constant occur- 
rence. 
♦ Ifc haa the deyiation of b, and ia double like it. No. 27 baa that of F. 
