6 NATURE 
[Marcu 4, 1915 


LETIERS LO PAE VEDIT OR 
[The Editor does not hold himself responsible for 
Opinions expressed by his correspondents. Neither 
can he undertake to return, or to correspond with 
the writers of, rejected manuscripts intended for 
this or any other part of Navurr. No notice is 
taken of anonymous communications.]| 
The Spectra of Hydrogen and Helium. 
In a letter to Nature of February 11, Prof. J. W. 
Nicholson discusses the recent interesting experiments 
of Mr. Evans on the spectrum emitted from a vacuum 
tube containing highly purified helium and subject 
to a heavy discharge. Evans found that the 4686 
series and the Pickering series can be obtained in a 
helium spectrum showing no trace of the ordinary 
hydrogen lines. These series were observed a few 
years ago by Prof. Fowler by sending a heavy dis- 
charge through a mixture of hydrogen and helium; 
previously they had been observed only in star spectra. 
In addition, Evans observed that under the same 
conditions as the 4686 series and the Pickering series 
a new series of lines appeared, which, with regard to 
position and intensity, could be united with the 
Pickering series into a single series of the same 
type as the 4686 series. The lines of the new series 
have wave-lengths very close to the hydrogen lines of 
the Balmer series. In his letter Prof. Nicholson 
expresses the opinion that Evans’s results cannot be 
used to discriminate between the different theoretical 
interpretations of the spectra in question, since the new 
series as well as the Pickering series can be deduced 
from the 4686 series on the general principle of com- 
bination of spectral lines. I cannot agree with this 
view, and should like here to state my reasons. 
According to the theory proposed by Rydberg, and 
generally accepted for a long time, all the lines in 
question are ascribed to hydrogen. This theory is 
based on a supposed analogy between the hydrogen 
spectrum and the spectra of the alkalies. The Balmer 
series, the Pickering series, and every second member 
of the 4686 series are considered as diffuse series, sharp 
series, and principal series. 
According to the theory proposed by the writer, 
the series spectrum of hydrogen is simply represented 
by the formula 
offal 
pee 
(as =) 
including only the Baimer series, the Ritz series in 
the ultra-red, and the series in the ultra-violet recently 
observed by Lyman. The other series are all ascribed 
to helium and represented by the formula : 
= K(4 1), 2 
ae ea ae ae 
including the 4686 series, the Pickering series, and 
Evans’s new series. K/’ is not exactly equal to K, 
though very nearly so, the theoretical value for the 
ratio K’/K being 100041. The theory rests upon a 
certain application of the quantum theory to the theory 
of the nucleus atom, and essentially involves the com- 
bination principle. Whiie the ordinary helium spec- 
trum is supposed to be emitted from a helium atom 
which has lost a single electron, spectrum (2) is 
emitted, according to the theory, from a helium atom 
which has lost both its electrons. If spark spectra 
result from the removal of two electrons from the 
atom, formula (2) should correspond to the spark 
spectrum of helium, while hydrogen, which is sup- 
posed to have only one electron in the atom, cannot 
give a spark spectrum. 
Since the numerical relations between the different 
NO. 2366, VOL. 95| 
(1) 
series claimed by the two theories are not exactly 
the same, it would be possible at once, as mentioned 
by Prof. Nicholson, to discriminate between the 
theories if the wave-lengths of the lines were exactly 
known. It seems that both the measurements of 
Fowler and Evans are in close agreement with 
formula (2), but that it would be difficult to reconcile 
Evans’s redetermination of the wave-length of the 
Pickering lines with Rydberg’s assumption that the 
Balmer series and the Pickering series have the same 
limit. Quite apart, however, from the question of 
the exact numerical relationship between the series, it 
seems that fundamental difficulties in Rydberg’s view 
are brought to light by the experiments of Fowler 
and Evans. Fowler’s observation that the 4686 series 
contains twice as many members as the series pre- 
dicted by Rydberg, and Evans’s observation of the 
new series accompanying the Pickering series, destroy 
completely the simple analogy with the alkali spectra 
which forms the basis of Rydberg’s theory. If, as 
Prof. Nicholson proposes, we look upon the Pickering 
series and the Evans series as a simple combination 
series deduced from the 4686 series, the contrast with 
the alkali spectra seems even more striking, since then 
apparently no sharp series is observed. 
hand, it will be noticed that this simple relation 
between the series is an immediate consequence of 
formula (2), and that Evans’s observation makes the 
representation of this formula more complete than 
before since only every second member of the series 
corresponding to m,=4 was previously observed. 
I must confess, however, that perhaps the strongest 
objection to Rydberg’s view seems to be the 
entirely different chemical conditions for the appear- 
ance of the Balmer series and the other series, so 
clearly shown by Evans’s experiments. This differ- 
ence has no analogy in the appearance of the diffuse 
series in the allali spectra. I am afraid that it may 
not seem reasonable to emphasise this objection so 
strongly, since it, of course, has been contemplated 
by spectroscopists ever since the theory was proposed, 
but it appears that the basis for the consideration of 
this question recently has entirely changed. Rydberg’s 
theory of the hydrogen spectrum offered not only an 
adequate representation of the evidence at hand at 
that time, but it was apparently the only way to 
bring the Pickering series into line with Rydberg’s 
own fundamental investigations on the general rela- 
tions between spectral series. However, Fowler’s 
recent important work on series in spark spectra has 
shown that these spectra obey laws of exactly the 
same character as the ordinary are spectra, with the 
only exception that the Rydberg constant is replaced 
by a constant four times as large. Consequently 
formula (2) corresponds to the general formula for 
series in spark spectra exactly as formula (1) 
corresponds to the general formula for series in arc 
spectra. It would therefore appear that at present 
there is scarcely sufficient theoretical evidence to 
justify us in disregarding the direct evidence as to 
the chemical origin of the lines given by Evans’s 
experiments. 
It may be of interest in this connection to mention 
that the above conclusions seem to be supported by 
spectroscopic evidence of an entirely different char- 
acter. Recently Dr. Rau (Sits. Ber. d. Phys. med. 
Ges. Wiirzburg, 1914) has made some _ interesting 
experiments on the minimum voltage necessary to 
produce spectral lines. He finds that about 30 volts 
are necessary to produce the lines of the ordinary 
helium spectrum, and that the voltage necessary to 
produce the different lines and series in this spectrum 
differ only by a few volts. On the other hand, he 
| finds that the lines of the Balmer series appear with 

On the other~ 
