118 
NATURE 

[JANUARY 27, 1923 
The Disappearing Gap in the Spectrum. 
By Prof. O. W. RicHarpson, F.R.S. 
Te 
pe Royal Institution seems a peculiarly fit place 
to deliver lectures on this subject, because it 
was while he was professor here 120 years ago that 
Thomas Young, the great advocate of the wave theory 
of light, showed how to estimate the wave-lengths of 
the different parts of the spectrum, and by so doing 
laid the foundations of spectroscopy as a quantitative 
science. His determinations of the wave-lengths in 
the visible spectrum were based on Newton’s observa- 
tions of the colours of thin plates. He also explained 
the principle of the diffraction grating, and by experi- 
ments based on the method of Newton’s rings he 
showed that the actinic or ultra-violet rays had shorter 
wave-lengths than those in the visible. The wave- 
lengths of the visible spectrum extend from a little 
below 4000 to a little above 7000 Angstrém 2 units, or, 
roughly, over about an octave. On the infra-red side 
we have, first, the invisible rays, often referred to as 
radiant heat, which contain the major part of the 
energy in the solar spectrum and a greater proportion 
still of the energy radiated from bodies at a lower tem- 
perature. Beyond these we have the long electro- 
magnetic waves of the type we are familiar with in 
wireless telegraphy. This side of the spectrum extends 
to waves of infinite length or of zero frequency. 
The gap in which we are interested is on the other 
side of the visible spectrum in the region of waves of 
shorter length or higher frequency. In 1801 Ritter 
showed that there was something beyond the violet 
end of the visible spectrum which blackened chloride 
of silver. In other words, there are ultra-violet rays 
which, as we should now say, are capable of ‘photo- 
chemical and photographic action. They also have 
other properties—they excite fluorescence in substances 
such as uranium glass, and they liberate electrons from 
the surface of a metal plate. They are, however, 
not very freely transmitted by glass; or, to put the 
matter more precisely, the ultra-violet spectrum which 
is transmitted by a glass prism spectroscope, does not 
extend very far beyond the visible limit. By substitut- 
ing quartz for glass in the spectroscope, and by other 
improvements, Stokes was able to make a very notable 
extension and to carry the limit to beyond 2000 
This made the ultra-violet extend over more than an 
octave, and measured in that way its extent had become 
greater than the whole of the visible spectrum. 
The limit to further extension was now found to be 
set by two things—(r) the absorption of quartz, which 
becomes fatal about 1850 A, and (2) the absorption 
of air, which also becomes prohibitive in the same 
neighbourhood. These difficulties were faced and 
overcome up to a certain point by Schumann, who 
constructed a fluorite spectroscope which he could 
operate, with all its adjustments, in an evacuated 
chamber. In this way he succeeded in pushing to the 
limit of transparency of fluorite, which is in the neigh- 
bourhood of 1250 A with good specimens. 
The limit to further development was set, and the 
* Substance of lectures delivered at the Royal Institution on May 
13 and 20 1922. 
2 y Angstrom unit (A)=1078 cm, 
NO. 2778, VOL. 111] 
| possible lines of advance narrowed down, by a very 
remarkable and important property of the radiation 
in this part of the spectrum, to wit, that every known 
material substance is practically completely opaque 
toit. I believe this high absorbability of the radiation 
to be due to the combined influence of two facts—(z) 
that the quantum of this radiation exceeds the ionisa- 
tion or radiation quantum of every atom, and (2) it 
does not exceed it by so much that there is any con- 
siderable chance of the radiation getting past the 
atom which, as it were, is set to trap it. We have 
precise evidence that absorption sets in as soon as, but 
‘not earlier than, the frequency at which the quantum 
of the impinging radiation exceeds the ionisation or 
radiation quantum of the atom. We also have con- 
siderable evidence, both theoretical and experimental, 
that the chance of absorption is greater when the two 
frequencies are comparable than when they are widely 
divergent in magnitude. These considerations exclude 
completely any apparatus of the type of the prism 
spectroscope, in which the radiation passes through 
considerable portions of matter such as the materials 
of the prisms and lenses. 
There is one spectroscopic apparatus which is free 
from this difficulty, namely, the concave grating 
invented by Rowland. In this device, if the slit, the 
grating, and the screen or photographic plate are all 
arranged to lie on a circle perpendicular to the rulings 
having a diameter equal to the radius of curvature of 
the grating, the spectrum is sharply focussed without 
using any lenses. The adaptation of the concave 
grating for use in this part of the spectrum is due to 
Lyman, whose vacuum grating spectroscope has only 
begun to bear the fruit which we may reasonably hope 
ultimately to gather from it. With this instrument, 
which I shall refer to more fully later, by 1913 Lyman 
had measured the wave-lengths of a large number of 
lines between the limits reached by Stokes (quartz) 
and Schumann (fluorite), and had also extended the 
known spectrum to the neighbourhood of goo A, which 
is the short wave limit of the most fundamental hydro- 
gen atom spectrum series, now known as the Lyman 
series. 
At that time, then, the spectrum was known to be 
continuous from wave-length infinity to wave-length 
goo A, or in terms of frequency from zero to 3°333 x 10% 
vibrations per second. It was also known that we had 
in the X-rays and the y-rays from radioactive sub- 
stances rays of still higher frequency and shorter wave- 
length. Prior to the discovery of the crystal diffraction 
phenomena the wave-lengths of X-rays had been 
ascertained roughly by photoelectric methods—a 
fact which seems generally to have been forgotten— 
but by 1913 they had been measured accurately by the 
Braggs and Moseley with the crystal spectrometer. 
Moseley’s measurements include such rays as the K- 
rays of aluminium, which are in the neighbourhood of 
8 A, and this was the longest X-ray wave then known. 
There was thus a gap from 8 A to goo A, or about seven 
octaves. This is the gap with which I propose to deal. 
I do not know that any systematic or very thorough 
attempt has been made to push the measurements of 

