The third room · Inside the lampChapter 14 of 21
In 1814 a Bavarian lens-maker spread sunlight more finely than anyone had before him, and found the rainbow crossed by hundreds of dark lines. For forty-five years nobody knew what they were.
Chapter 14 · Inside the lamp
The Line
Why each element shines in colours of its own, and what those colours revealed about atoms, and about the stars.
Look first
Salt in a flame
A clean gas flame is pale blue and gives little light. Sprinkle table salt into it and it flares a strong yellow. That yellow is not a glow of the kind you met in the last chapter. Spread it with a prism, and where a hot wire gives the whole band from red to violet, the salted flame gives almost nothing but one narrow line, at 589 nanometres.
The line belongs to sodium, one of the two elements in salt. Lithium gives a red line instead, at 671 nanometres. Hydrogen, glowing in a glass tube with a current through it, gives four. Every element has a set of its own, as particular to it as a signature, and no temperature changes where the lines fall.
The lines are drawn where NIST measures them. Their widths and brightnesses are not to scale.
Commit to a prediction
A flame in front of a lamp
Take the salted flame, at about 2,000 kelvin, and its yellow line at 589 nanometres. Now set a white-hot lamp behind it, a filament at 3,000 kelvin, far brighter than the flame. Look through the flame at the lamp, and spread the light into a spectrum.
What do you see at 589 nanometres?
Four nanometres of spectrum around sodium’s pair of lines. Each strip is drawn to its own brightest point: the flame alone is far fainter than the lamp.
The lines in the flame’s light and the gaps in the lamp’s fall in exactly the same places. In 1859 Gustav Kirchhoff saw what that meant for Fraunhofer’s dark lines. One of the strongest, which Fraunhofer had labelled D, sits at 589 nanometres. Sodium, he concluded, is in the atmosphere of the Sun: a cooler gas in front of a hotter glow.
Why the first guesses made sense
It adds and it takes away
If you chose the brighter line, you were right that the flame adds its yellow light. If you chose no change, you were right that the flame is transparent to nearly all of the lamp’s light. What both leave out is that sodium atoms take in light at exactly the wavelength at which they give it out, and at no other.
So at 589 nanometres the flame does two things at once. It removes some of the lamp’s light, and it adds some of its own. Which wins depends on one thing only: which is hotter. Before a hotter lamp the atoms take more than they give back, and the line is dark. Before a cooler one, or with nothing behind at all, the line is bright. At equal temperatures the two balance exactly and the line vanishes, like the cups in the kiln of chapter 13. It is the same law.
Why a line at all
A ladder with fixed rungs
A hot wire gives every colour because its atoms are packed together and jostle one another in every possible way. The atoms in a flame are far apart, each on its own. And an atom on its own, it turns out, can hold only certain amounts of energy, like a ladder that can be stood on only at its rungs.
When an atom drops from one rung to a lower one, the energy it loses leaves as a single lump of light. The lump’s size is the gap between the rungs, and its size sets its colour: sodium’s yellow is a gap of 2.1 electron volts. An atom on the lower rung can climb back by taking in a lump of exactly that size, and no other will do. That is why the bright line and the dark line coincide.
Planck was forced to his lumps by the shape of a glow, and distrusted them. Here they can be seen with the eye: one sharp colour for every pair of rungs.
The bench
Bright line, dark line
The flame stays at 2,000 kelvin. You choose how much sodium is in it, and how hot a glowing body stands behind it.
Sodium’s lines, 587 to 591.5 nm
Things to try
- Slide the glow behind from cool to hot. At what temperature do the lines disappear?
- With a hot glow behind, add more and more salt. Do the lines ever go completely black?
- Take the glow away. How bright can the flame’s line become, however much salt you add?
What you can now say
Every element gives out light, and takes it in, at sharp wavelengths of its own, because its atoms can hold only certain energies and each line is a jump between two of them. A cooler gas in front of a hotter glow leaves dark lines. Alone, it gives bright ones.
The lines are the same in a flame, in a laboratory tube and in a star. That is how anyone knows what a star is made of.
For the determined reader: the formulas, the lines, and the limits
Light through a layer of gas
A layer of gas at temperature Tg stands in front of a body glowing at Tb. At some wavelength the gas has optical depth τ: it lets through a fraction e−τ of the light from behind. If collisions keep its atoms in step with its temperature, it also adds its own glow, and what comes out is
I = B(Tb) e−τ + B(Tg) (1 − e−τ)
where B is the black-body curve of chapter 13. Away from a line τ is nearly zero, and the light from behind passes untouched. On the line, the result lies between B(Tb) and B(Tg), and approaches B(Tg) as the gas thickens. However much salt is added, the flame’s line can never outshine a black body at the flame’s own temperature.
The lines drawn in this chapter
| Element | Lines, in air | Seen as |
|---|---|---|
| Sodium | 588.995 and 589.592 nm | yellow; Fraunhofer’s D |
| Lithium | 670.78 nm | red |
| Hydrogen | 656.279, 486.135, 434.047, 410.173 nm | red, blue-green, violet, violet |
| Helium | 587.562 nm | yellow, beside sodium’s |
| Calcium, with one electron gone | 393.366 and 396.847 nm | violet; Fraunhofer’s K and H |
Where this model stops
A real flame’s lines are a few thousandths of a nanometre wide. They are drawn here about twenty times wider, as a modest spectroscope would blur them. The model takes the gas to be at one temperature, with collisions frequent enough to keep its atoms in step with that temperature. In a thin gas, such as the outer atmosphere of the Sun, that fails, and an atom that has taken in light may simply send it out again in another direction. The line is still dark, since the light is still removed from the beam, but its depth is no longer set by a temperature alone.
A case you have not seen
Hydrogen’s next line
Hydrogen is the simplest atom: one electron. Its visible light is four lines, at 656.3, 486.1, 434.0 and 410.2 nanometres. In 1885 those four numbers were all that was firmly known, and nobody could say why they had those values and not others.
Is there a fifth line, and if so, where?
Above: hydrogen’s lines, computed from the constants of nature and marked where NIST measures them. Below: the rungs of the hydrogen atom. Every line in the picture above is a fall to the second rung.
The pattern was found by Johann Balmer, who taught mathematics at a girls’ school in Basel, in 1885, when he was sixty. He had no theory, only a talent for numbers: each wavelength is 364.56 nanometres multiplied by n²/(n² − 4), with n = 3, 4, 5, 6. His formula gave a fifth line at 397 and an end to the series at 364.6, and both were there.
Nobody knew why until 1913, when Niels Bohr proposed the ladder. If the rungs of hydrogen lie at 13.6 electron volts divided by 1, 4, 9, 16 and so on below the top, then every fall to the second rung gives one of Balmer’s lines, and the 13.6 can be worked out from the charge and mass of the electron and Planck’s constant. Computed that way, with nothing adjusted, the red line comes out at 656.288 nanometres. It is measured at 656.279.
Bohr’s paper adds a remark that reaches ahead to the stars. Only twelve of Balmer’s lines had been seen in laboratory tubes, he noted, while thirty-three were seen in the light of stars. That is what the ladder expects: an atom on a high rung is very large, and survives only where the gas is thin.
One more case
The Sun’s faint hydrogen
Look again at the spectrum of the Sun at the top of this chapter. The two heaviest lines, in the violet, belong to calcium. Hydrogen’s lines are there too, in the red and the blue-green, but they are modest. In a hotter star such as Sirius it is the other way round: hydrogen’s lines are the strongest in the whole spectrum.
What does that tell you about the Sun?
Of all the hydrogen in a star’s atmosphere, the share that sits on the second rung and can absorb the visible lines. Each step of the scale is a factor of ten.
Hydrogen’s visible lines all start from the second rung, which lies 10.2 electron volts above the first. At the Sun’s temperature only about one hydrogen atom in 200 million is there. The rest sit on the bottom rung, where they can absorb only ultraviolet light. Near 10,000 kelvin the share is nearly two thousand times greater. Hotter still, the atoms lose their electrons altogether, and the lines fade again.
Calcium is the opposite case. In the Sun nearly every calcium atom has lost one electron, and that is the form which makes the two violet lines, starting from its lowest rung. So a rare element writes the heaviest lines in the spectrum, and the commonest one writes faint ones.
Cecilia Payne worked this out in her doctoral thesis of 1925, and concluded that the stars are made overwhelmingly of hydrogen and helium. Henry Norris Russell, one of the leading astronomers of the day, urged her not to say so, and she added a sentence calling her own result “almost certainly not real”. Four years later Russell reached the same conclusion by another route.
Reading the lines
William Wollaston noticed a few dark gaps in the Sun’s spectrum in 1802. In 1814 Joseph von Fraunhofer, a glassmaker’s son who had become a maker of lenses and optical glass, found them again, mapped more than 570, and gave the strongest the letters A to K that they still carry. He had no idea what they were.
The answer came forty-five years later, in Heidelberg. Robert Bunsen had helped to develop a burner with a clean, nearly colourless flame, and with the physicist Gustav Kirchhoff he used it to study the light of heated elements. The test was sensitive beyond anything chemistry had. They exploded three milligrams of a sodium salt in the far corner of their laboratory, and within minutes the flame at the other side of the room turned yellow. In 1859 Kirchhoff showed that the Sun contains sodium. In 1860 the two of them found lines that belonged to no known element in a mineral water, and so discovered caesium, the first element found with a spectroscope and named for the blue of its lines. Rubidium followed in 1861. Their paper of 1860 says where the method leads: it “opens to chemical research a hitherto completely closed region extending far beyond the limits of the earth and even of the solar system”.
During the eclipse of 18 August 1868 a yellow line was seen in the light from the edge of the Sun, close to sodium’s but not on it. Norman Lockyer proposed that it belonged to an element unknown on Earth, and named it helium, after the Sun. It was found on Earth in 1895.
Where this explanation stops
Why are there rungs?
Bohr’s ladder gives hydrogen’s lines and fails for nearly every other atom. It also does not say why an atom should have rungs at all. That took about a dozen more years and a new mechanics, in which the electron in an atom is itself a kind of standing wave, and only certain waves fit. The rungs are the notes of the atom.
An atom gives out its light as one whole lump, all at once. Yet that same light goes on to pass through two slits and make the fringes of chapter 3. How does a single lump do what only a wave can do?
End of chapter 14
Added to your notebook
Every element gives out light, and takes it in, at sharp wavelengths of its own, because its atoms can hold only certain energies. A cooler gas in front of a hotter glow leaves dark lines. Alone, it gives bright ones.
One atom, one lump of light. How does a single lump make the fringes of a wave?
Chapter 15 · Every Path