978 
EEV. T. E. EOBINSON ON SPECTEA OF ELECTEIC LIOHT. 
the ordinary character of which is bn. When examined with the S^C prisms it is triple, 
the first being very narrow but bright, with a slight tinge of orange, the second and 
third pure yellow, the second broader than the third, and both sharply defined. 
Nos. 26 to 29 form a very conspicuous group. The first, ^ (which begins the green), is 
marked by the very obscure interval which separates it from the others ; it as well as 
g" are double, and the S^C prisms show many fine lines in their neighbourhood. With 
iron, copper, and in many of the CO spectra, the number of these is very great ; it may 
be that the bright green from this to No. 44 is composed of such lines too close to be 
resolved ; but I think it is continuous. 
No. 36 is nearly in the place of E. 
No. 38 ;j is very conspicuous on this bright ground; it is often intensely brilliant, 
sometimes more so than and it occurs more frequently. No. 37 seems to belong to it; 
with the S^C prisms, and sometimes with the glass one, it is double of two equal. Its 
brightness is scarcely developed in O, C.P., very much in N, but quite as much in 0, E., 
as in N, E. CO, E. exceeds them both, though it has only 2 *s in C.P. With platinum, 
CO, C.P., it is replaced by a bright bluish-green band of singular appearance ; this by 
careful focusing shows eleven or twelve fine lines, of which No. 37 is the first and No. 42 
the last. In this spectrum the group g, g" is also replaced by some twenty-five very fine 
lines from No. 23 to No. 36. The place of ri is identical with the centre of the double 
line of h. 
No. 44 0^ with its companions 42 and 43, is very common, though the latter, 43 
especially, are hard to see in the blinding glare of the first, which is often most intense. 
They are always seen with S‘^C prisms, which also show 6 double (as Duboscq does occa- 
sionally). The components are generally, but not always, equal*. Though frequent in 
E. it is seldom bright. In C.P., N brings it out best; it is wanting in 8 of O, 8 of H, 
5 of CO ; and of its 50 * O has but 4, H I, and CO 3 ; one belongs to mercury in 
mercury vapour. It ends the green. 
No. 49 is remarkable, not only from its being in the place of F, but also from its 
being (or being in) the characteristic blue band of H, C.P. In the other gases and 
vapours it is scarcely ever a #, and often faint ; here it is of great brightness and great 
breadth, often 6' or 7'. It has a cloudy but irresolvable look, and its edges are not 
sharply defined. Even with 2 S^C prisms I cannot resolve it; and it gives me the 
impression that it consists of light whose wave-length varies continuously. Its mean by 
22H, C.P., is 33° 4I'T8, corresponding to a wave length 1798. It is singular that the 
N, C.P., spectrum of tellurium has a band possessing this peculiar type so decidedly that 
one who saw it without knowing its origin would undoubtingly assert the presence of H ; 
and my first idea was that this gas must be a component of the metal. If, however, 
* In this, as in similar instances, I suspect that an unequal development of the components may disturb 
the measures. I may at the same time notice another cause of disturbance, the flicker of the light. It 
might be supposed that the collimator must give an object absolutely fixed, but this is not the case. The 
discharge being narrower than the slit and at some distance from it, and the want of perfect achi’omatism 
in the object-glass, are capable of producing considerable unsteadiness. 
