Like other chemists of the day, Bunsen had convinced himself that color is key to the photochemical identification of elements—that each element, incinerated in a flame, blazes with a unique color that, if properly characterized, establishes the element’s presence. But how to define that telltale tinge, how to distinguish this shade of yellow from that, how to describe a complex color formed from an amalgam of hues? For more than a year, Bunsen and a student from England, Rowlandson Cartmell, had exhaustively examined the colored emission of flames. Their goal was to construct a series of glass and chemical filters to mask colors released by known impurities such as sodium, thereby disclosing the true characteristic color of a sample. Gustav Kirchhoff was brought in to gauge precisely which wavelengths of light were blocked by the various filters. How he accomplished this is undocumented, but he probably used a spectroscope to assess each filter’s selective absorption of sunlight. Indeed, only a few months later, in mid-1859, Kirchhoff took up a parallel study of the optical properties of crystals. Although his measurement apparatus is not pictured, Kirchhoff’s detailed description is of a sophisticated, high-magnification spectroscope.
Around the same time, Bunsen admitted in a letter to Henry Roscoe that he had squandered his time on color-filter chemical analysis. Yes, the method did work to a limited degree for select substances. But mulling over its expanded use, Bunsen recognized the folly of developing the requisite multitude of filters. Yet he clung to his faith that chemical elements could be characterized by their light.
Just as paint incorporates individual pigments, the light of incandescent materials is a mixture of discrete colors. However, the human eye maps images based on the spatial distribution of the amalgamated light from a source, and is poor at deducing the presence—much less the relative intensities—of specific color components. Thus, two glowing elements might differ markedly in their color constituents, offering a perceptible difference in the character of their emitted light. Another pair of elements might differ only slightly in their luminous makeup, presenting no discernible difference to the eye. Even an element that appeared orange in a flame might harbor subtle—and unseen—shades of blue or violet. In essence, Bunsen lacked the means to assay the pigments dissolved in the paint. And the obstacle, he came to realize, lay not in his familiar domain of chemistry, but in the realm of physics.
Perhaps it was during one of their late-summer ambles along Heidelberg’s Philosopher’s Walk that Bunsen asked his friend Kirchhoff for advice on his experimental conundrum: How does one definitively parse the colors within light? Kirchhoff would have had a ready answer; the apparatus he had only recently used to study crystals contained precisely the instrument Bunsen needed—a spectroscope. Kirchhoff reasoned that if each incandescent element emits light of a characteristic color, as Bunsen asserted, then the spectrum of that light must necessarily be unique to that element. The spectroscope would disperse the intertwined colors of the flame into an ordered regiment of spectral lines, whose pattern corresponds to a given element.
In the early autumn of 1859, Bunsen’s and Kirchhoff’s experimental goals converged: one’s dream of practical photochemical analysis melded with the other’s research into the optical properties of matter. The two scientists retreated to their respective laboratories: Bunsen to his gleaming Chemical Institute, where he prepared highly purified samples of chemical compounds; Kirchhoff to a back room in the dilapidated physics building several blocks away, where he adapted his spectroscopic apparatus to the stringent needs of the new project.
The spectroscope’s optical components were housed in a squat, trapezium-shaped enclosure, elevated a few inches above a wooden base. At the center of this homely “cigar box” was a prism-shaped bottle containing carbon disulphide, a fluid with almost twice the light-dispersing power of expensive flint glass. The box’s side walls carried a pair of small, fixed telescopes: one to guide a sliver of incident light into the prism, the other to view the emergent spectrum. The first of these telescopes was inserted backwards: its objective lens sat inside the box near the prism, while its focus-end—normally an eyepiece, but here a brass-edged entry slit—protruded horizontally from the box. This telescope served as a collimator, taking incoming light rays, which diverge upon passage through a slit, and rendering them parallel before they reach the prism. (Developed in the 1830s by English optician William Simms, the collimator maximizes a spectroscope’s resolving power—its ability to render fine details in a spectrum.) The second telescope, on the opposite side of the box, was fitted with a vertical sighting wire to assist in placing the various spectral features dead-center in the field of view.
Bunsen’s chemical samples, each cradled in a tiny loop of platinum wire, were successively incinerated in a gas burner placed before the spectroscope’s entry slit. Mounted on a swivel, the central prism could be rotated azimuthally to bring different parts of the spectrum into view. The swivel, in turn, held a vertical rod, which poked out the bottom of the enclosure and bore a small mirror. The mirror allowed Kirchhoff to sight, through an external telescope, the reflection of a finely graduated scale several feet away. By axial adjustment of the prism, any spectral line could be brought into coincidence with the vertical wire in the viewing eyepiece; the associated scale reading quantified the line’s position.
The original spectroscope used by Gustav Kirchhoff and Robert Bunsen to explore the spectra of chemical elements at Heidelberg in 1859.
By April 1860, Kirchhoff had subjected several of Bunsen’s chemical specimens to spectroscopic analysis. The acuity of their home-brewed spectroscope was sufficient to reveal faint emission lines never before seen. Kirchhoff was easily able to distinguish among the spectral-line patterns of various elements, including sodium, lithium, potassium, barium, strontium, and calcium. He compared elemental spectra produced in a variety of flames and in electrical spark gaps to prove that line patterns are innate to matter itself and not affected by either the mode or temperature of incandescence.
Both men acknowledged that Bunsen’s impeccable laboratory technique was crucial to their enterprise. Bunsen was vigilant about eliminating impurities—especially the ubiquitous sodium—that would have otherwise muddled the interpretation of the spectra. In one test, Bunsen burned several milligrams of salt in the corner of the laboratory opposite the spectroscope. Within a few minutes, sodium lines appeared in the spectrum of a previously featureless gas flame, the sodium particles having wafted across the room. Bunsen reported that the combustion of as little as one three-millionth of a milligram of sodium generates a perceptible line. (He added that airborne sodium might play a role, for good or for ill, in the spread of contagious disease.) In one stroke, Bunsen’s test cautioned would-be spectrochemists against airborne contaminants, while affirming the exquisite sensitivity of the spectroscope itself.
That a seasoned “wet-chemist” like Bunsen would endorse spectroscopic analysis lent special credence to the new process. He and Kirchhoff predicted the discovery of new elements by optical means, “for if bodies should exist in nature so sparingly diffused that the analytical methods hitherto applicable have not succeeded in detecting, or separating them, it is very possible that their presence may be revealed by a simple examination of the spectra produced by flames.”
Barely a year later, in June 1861, Bunsen and Kirchhoff fulfilled their own prediction with the spectroscopic discovery of two new elements: cesium and rubidium. To obtain a sufficient quantity of cesium for spectroscopic analysis, Bunsen had refined forty tons of mineral water from the springs at Dürkheim. The “mother-liquor,” as he called it, was chemically stripped of known elements until the resulting concentrate displayed a pair of previously unseen blue emission lines. (The name cesium, or caesium, derives from the Latin caesius, for bluish gray.) A similar effort was mounted against 180 kilograms of the mineral lepidolite before the distinctive red line of rubidium rose to visibility. The dual discovery sent chemists scurrying to their laboratories to set up their own spectros
copes. Before decade’s end, three more elements were discovered by spectroscopic means.
One evening, Bunsen recalled, he and Kirchhoff were working late in the laboratory, when they noticed a golden glow on the horizon. A fire was raging in the neighboring city of Mannheim, toward the northwest. Ever curious, they moved their spectroscope to the window and peered at the dispersed light. By now, both were intimately familiar with the spectral patterns of chemical elements. From the brilliant duo of green lines, which they had previously dubbed alpha and beta, they surmised that the fire had ignited compounds containing barium. And astride the barium pair, they recognized the distinctive eight-line fingerprint of strontium: six reds and an orange to one side, a lone blue to the other. If the chemical constituents of a blaze could be discerned at ten miles, Bunsen mused aloud to Kirchhoff, might not future spectroscopists just as reliably assay the remote stars?
As far as Kirchhoff was concerned, the future had already arrived. During the first weeks of his collaboration with Bunsen, he modified the spectroscopic apparatus to allow direct comparison of laboratory-generated spectra with that of the Sun. In a hastily composed announcement to the Berlin Academy, dated October 20, 1859, Kirchhoff alludes to his and Bunsen’s unfolding success identifying chemical substances from their flame spectra. Yet he declines to elaborate, and shifts abruptly to the matter at hand: “I made some observations which disclose an unexpected explanation of the origin of Fraunhofer’s lines, and authorize conclusion therefrom respecting the material constitution of the atmosphere of the sun, and perhaps also of the brighter fixed stars.”
Using a clock-driven plane mirror (a “heliostat”), Kirchhoff had projected a sunbeam through a salt flame before it entered the spectroscope. Manipulating a sequential pair of calcite crystals, he increased or decreased the intensity of sunlight illuminating the flame. When the solar illumination was low—that is, when the flame was not backlit by the Sun—the lines of sodium appeared bright, as in a typical flame spectrum. However, when the solar backlighting was intensified, the bright lines transformed into dark lines. Kirchhoff reasoned that, while incandescent sodium always throws off its characteristic rays, these lines of emission were now seen in silhouette against the more brilliant Sun. Indeed, when Kirchhoff blocked the sunlight, the sodium lines reverted to their bright form. Moving the salt flame into and out of the solar beam, he found that the Sun’s D lines were darker when the flame was interposed than when it was not. That is, the flame’s radiance did not “fill in” the dark D lines, as he had expected, but reinforced the absorption of these wavelengths of light. Functionally, the flame was a far-flung extension of the solar atmosphere, both of them acting in concert on the light emanating from the Sun’s interior. More surprising, an interposed lithium flame created a dark line in the Sun’s spectrum where no such line had existed before; two superimposed luminous sources—the Sun plus the flame—somehow resulted in a selective diminution of light.
To reaffirm the evidence gleaned from nature, Kirchhoff decided to create Fraunhofer lines using an artificial sun. He acquired a Drummond lamp—the original “limelight” of theatrical fame—and swapped out the Bunsen burner for a less-intense alcohol flame. Not only does a Drummond lamp burn hotter than alcohol, its spectrum is virtually continuous like the Sun’s. (The lamp’s inevitable sodium-contaminant emission disappears after brief use.) As Kirchhoff interposed the salt-laced flame between the lamp and the spectroscope, he saw the bright D doublet go dark, as in the solar spectrum. He had artificially impressed the telltale sodium lines upon the spectrum of a lamp that itself contains no sodium. His conclusion was that sodium in a flame emits characteristic wavelengths, but also absorbs the same wavelengths from light passing through it. Kirchhoff asserts, in his 1859 announcement, “that the dark lines of the solar spectrum . . . exist in consequence of the presence, in the incandescent atmosphere of the sun, of those substances which in the spectrum of a flame produce bright lines in the same place.” Fraunhofer’s D lines, manifested in emission or absorption, mark the presence of sodium in a laboratory flame, in the Sun, or in a star.
In the months that followed, Kirchhoff developed a physical theory to account for the origin of spectral lines. Analogizing to the natural flow of heat from a region of high temperature toward that of low temperature, he deduced that a body with a propensity to emit light at a given wavelength must have an equal propensity to absorb light at that wavelength. Thus, the luminous emission of a hot, diffuse gas takes the form of bright lines when viewed directly, but dark lines when viewed against an incandescent body of higher temperature. (In fact, the bright-dark reversal of sodium lines was observed and similarly explained in 1849 by French physicist J. B. L. Foucault, who superimposed solar and electric-arc spectra. Kirchhoff was unaware of Foucault’s work.)
Kirchhoff’s radiation theory explains spectral-line formation to the degree that Freudian theory explains the psychological manifestations of brain function. They are functional models of phenomena whose root cause, in their respective eras, were unknowable. Thus, Kirchhoff says nothing about the atomic-level phenomenon that gives rise to the emission or absorption of light by matter, nor could he: the structure of the atom was as yet unknown. However, the prevailing idea that the atom is indivisible—that it represents the fundamental iota of matter—was shaken by the sheer number and variety of spectral lines associated with any given element. How could a unitary particle give off light at a multitude of discrete wavelengths? Scientists suspected that spectral lines reflected something about the form and behavior of the atom, but they were unable to decode the clues. In the absence of a realistic atomic model, which would not come until the twentieth century, physicists proposed scenarios of spectral-line production that typically involved complex vibrations of the atom.
Spectra of the Sun, Sirius, and several chemical elements. Note the coincidence of the dark D-lines in the stellar spectra with their bright-line counterparts in the laboratory spectrum of sodium (Na).
Having placed spectral-line production on a theoretical footing—at least to his own satisfaction—Kirchhoff, in 1861, stated explicitly what he had only intimated in his prior communication to the Berlin Academy:
The sun possesses an incandescent, gaseous atmosphere, which surrounds a solid nucleus having a still higher temperature. If we could see the spectrum of the solar atmosphere, we should see in it the bright bands characteristic of the metals contained in the atmosphere. . . . The more intense luminosity of the sun’s solid body, however, does not permit the spectrum of its atmosphere to appear; it reverses it . . . so that instead of the bright lines which the spectrum of the atmosphere by itself would show, dark lines are produced. Thus, we do not see the spectrum of the solar atmosphere, but we see a negative image of it. This, however, serves equally well to determine with certainty the presence of those metals which occur in the sun’s atmosphere.
Kirchhoff’s spectroscopically founded assertion that the solar interior is hotter than its luminous envelope marked a break with decades of speculative tradition. The famed observer William Herschel had claimed that the visible Sun is actually a hot, opaque cloud-layer concealing a central world whose surface is temperate—and likely inhabited. Sunspots were considered to be fleeting breaks in the clouds through which one might glimpse the hospitable regions below. François Arago, director of the Paris Observatory and early champion of photography, assured an 1840s audience that “if anyone asked me if the Sun could be inhabited by a civilization like ours, I would not hesitate to say, yes.” As late as 1854, English physicist David Brewster maintained that “we approach the question of the habitability of the sun, with the certain knowledge that the sun is not a red-hot globe, but that its nucleus is a solid opaque mass receiving very little light and heat from its luminous atmosphere.”
English astronomer J. Norman Lockyer declared in 1881 that Kirchhoff’s spectroscopic observations “at once destroyed, at a blow, the idea . . . that the sun was a cool habitable globe, with t
rees, and flowers, and vales, and everything such as we know of here. If the atmosphere were in a state of sufficient incandescence to give these [spectral-line] phenomena it was absolutely impossible that anything below that atmosphere should not be at the same time at a high temperature.”
Kirchhoff’s 1861 report presents the first evidence-based solar model fully informed by the maturing physics and chemistry of the mid-nineteenth century. Crude as it may read in retrospect, the model’s underlying implication was profound: The spectroscope had effectively projected the scientist’s critical faculties almost a hundred million miles, into the very body of the Sun. A “victory over space,” trumpeted physicist Arthur Schuster in a review for Popular Science Monthly. Astronomers would never look at the Sun the same way: It was, henceforth, a physical body—a star—whose structure and luminance arise from, and are ever subject to, physical law. Nor could scientists fail to recognize the immense potential of spectroscopy, both in the cosmic realm and in the laboratory.
By early 1860, Bunsen and Kirchhoff had acquired a pair of improved spectroscopes from the firm of C. A. von Steinheil in Munich. The newer instrument featured a sequential quartet of flint-glass prisms, the latter trio each dispersing the spectrum transmitted by its predecessor. No other spectroscope in the world could display spectral features with comparable resolution. A small reflecting prism grafted to the entry slit allowed the spectra of two sources to be seen simultaneously in the eyepiece, one directly above the other. This feature was put to immediate use to better compare solar and laboratory-generated spectra. (By this time, Bunsen had provided Kirchhoff with samples of thirty-two elements that could be set alight in a flame or spark.)
Starlight Detectives Page 20