(Background, above) The Tarantula Nebula, known also as 30 Doradus, is the largest and richest stellar nursery in the known universe. Stretching for several hundred parsecs, its interstellar clouds are illuminated by the strong ultraviolet radiation from the young blue-white stars shining amidst the darkness of the nighttime sky.

(Inset A) The Cone Nebula is part of a large and messy complex of nebulosity in the constellation Monoceros. The cone itself seems to be a clump of dense clouds of dust, thick enough to keep out, for now, the penetrating radiation surrounding it.

(Inset B) The big, overexposed star at left is Antares--a red supergiant star shedding gas and particles from its swollen surface. The different colors here result from the scattering of Antares' light among dusty interstellar grains of many sizes.

(Inset C) A better example of a reflection nebula is this bluish molecular cloud. It has the innocuous name CG4, and is about 300 pc away. The only part of the nebula that is aglow itself is the reddish matter near the center; all the rest of the blue color is caused by scattered light, much like that causing our blue sky on Earth.

(Inset D) This emission nebula, IC2944, is a parsec-sized cloud of mostly hydrogen gas illuminated by a few very hot stars. Embedded within the nebula are several tiny dark clouds, known as Bok globules, that might now be contracting to become more stars.


Studying this chapter will enable you to:

Summarize the composition and physical properties of the interstellar medium.

Describe the characteristics of emission nebulae, and explain their significance in the life cycle of stars.

Discuss the nature of dark interstellar clouds.

Specify the radio techniques used to probe the nature of interstellar matter.

Discuss the nature and significance of interstellar molecules.

In addition to stars and planets, our Galaxy also harbors matter throughout the invisible regions of interstellar space, in the dark voids between the stars. The density of this interstellar matter is extremely low--approximately a trillion trillion times less dense than matter in either stars or planets, far more tenuous than the best vacuum attainable on Earth. Only because the volume of interstellar space is so vast does its mass amount to anything at all. So why bother to study this near-perfect vacuum? We do so for three important reasons. First, there is as much mass in the "voids" between the stars as there is in the stars themselves. Second, interstellar zspace is the region out of which new stars are born. Third, it is the region into which some old stars explode at death. It is one of the most significant crossroads through which matter passes in our universe.

Back Figure 18.1 shows a large region of space, a much greater expanse of universal real estate than anything we have studied thus far. The bright regions are congregations of innumerable stars, some of whose properties we have just studied in Chapter 17. However, the dark areas are not simply "holes" in the stellar distribution. They are regions of space that obscure or extinguish light from the stars beyond. These regions house interstellar matter, consisting of great clouds of gas and dust. Their very darkness means that they cannot be easily studied by the optical methods used for stellar matter. There is, quite simply, very little to see!

Figure 18.1 A wide-angle photograph of a great swath of space, showing regions of brightness (vast fields of stars) as well as regions of darkness (obscuring interstellar matter). The field is roughly 30° across.

Figure 18.1 shows that the dark interstellar matter is rather patchy. It is spread very irregularly through space. In some directions, the obscuring matter is largely absent, and astronomers can study objects literally billions of parsecs from Earth. In other directions, the obscuration is moderate, prohibiting visual observation of objects beyond more than a few thousand parsecs, but still allowing us to study nearby stars. Still other regions are so heavily obscured that starlight from even relatively nearby stars is completely absorbed before reaching the Earth.


The matter between the stars is called the interstellar medium. It is made up of two components--gas and dust--intermixed throughout all of space.

Interstellar gas is made up mainly of individual atoms, of average size 10-10 m (1A) or so. The gas also contains some small molecules, no larger than about 10-9 m across. Regions with such small particles are transparent to nearly all types of radiation, including ultraviolet, visible, infrared, and radio waves. Apart from numerous narrow atomic and molecular absorption lines, gas alone does not block radiation to any great extent.

Interstellar dust is more complex. It consists of clumps of atoms and molecules--not unlike chalk dust and the microscopic particles that make up smoke, soot, or fog. Light from distant stars cannot penetrate the densest accumulations of interstellar dust any more than a car's headlights can illuminate roadside objects in a thick Earth-bound fog. Comparisons of how starlight is diminished in interstellar space with the scattering of light in terrestrial fog indicate that the typical size of an interstellar dust particle--or dust grain--is about 10-7 m. The grains are thus comparable in size to the wavelength of visible light, and about 1000 times larger than interstellar gas particles.

The ability of a particle to scatter a beam of light depends on (1) the size of the particle and (2) the wavelength of the radiation involved. (See the More Precisely feature on p. 152.) As a rule of thumb, only particles with diameters comparable to or larger than the wavelength can significantly influence the beam. Because the wavelength of radio waves greatly exceeds the size of the dust grains, dusty interstellar regions are completely transparent to long-wavelength radio radiation. In other words, radio waves pass through dusty regions unimpeded. These regions are also partially transparent to infrared radiation. Conversely, interstellar dust is very effective at blocking short-wavelength optical, ultraviolet, and X-ray radiation. This general dimming of starlight by interstellar matter is called extinction.

Because the interstellar medium is more opaque to short-wavelength radiation of than to radiation of longer wavelengths, light from distant stars is preferentially robbed of its higher-frequency ("blue") components. Hence stars also tend to appear redder than they really are, an effect known as reddening. It is similar to the process that produces spectacular red sunsets on Earth.

As illustrated in Figure 18.2, extinction and reddening change a star's apparent brightness and color, but they have no effect on its spectral type. Absorption lines in the star's spectrum are largely unaffected by interstellar dust. Astronomers can use this fact to study the interstellar medium. By determining a main-sequence star's spectral type, astronomers first learn its true luminosity and color. They can then measure the degree to which the starlight has been affected by extinction and reddening en route to Earth. This, in turn, allows them to estimate both the numbers and the sizes of interstellar dust particles along the line of sight to the star. By repeating these measurements for stars in many different directions and at many different distances from the Earth, a picture of the distribution and properties of the interstellar medium in the surrounding solar neighborhood can be built up.

Figure 18.2 Starlight passing through a dusty region of space is both dimmed and reddened, but spectral lines are still recognizable in the light that reaches Earth.


The temperature of the interstellar gas and dust ranges from a few kelvins to a few hundred kelvins, depending on its proximity to a star or some other source of radiation. Generally, we can take 100 K as an average temperature of a typical dark region of interstellar space. Compare this with 273 K, at which water freezes, and 0 K, at which atomic and molecular motions all but cease. Interstellar space is very cold.

The density of interstellar matter is extremely low. It averages roughly 106 atoms per cubic meter--just 1 atom per cubic centimeter--but densities as great as 109 atoms/m3 (1000 atoms/cm3) and as small as 104 atoms/m3 have been found. Matter of this low density is far more tenuous than the best vacuum--about 1010 molecules/m3--that we can make in laboratories here on Earth. Interstellar matter is about a trillion trillion times less dense than water. Interstellar dust is even rarer than interstellar gas. On average, there are only about 10-6 dust particles per cubic meter--that is, 1000 per cubic kilometer. Interstellar space is populated with gas so thin that harvesting all the matter in an interstellar region the size of Earth would yield barely enough matter to make a pair of dice.

How can such fantastically sparse matter diminish light radiation so effectively? The key is size--interstellar space is vast. The typical distance between stars (1 pc or so in the vicinity of the Sun) is much, much greater than the typical size of the stars themselves (around 10-7 pc). Stellar and planetary sizes pale in comparison to the vastness of interstellar space. Thus, matter can accumulate, regardless of how thinly spread. For example, an imaginary cylinder 1 m2 in cross section and extending from Earth to Alpha Centauri would contain more than 10 billion dust particles. Over huge distances, dust particles accumulate slowly but surely, to the point at which they can effectively block visible light and other short-wavelength radiation. Even though the density of matter there is so low, interstellar space in the vicinity of the Sun contains about as much mass as exists in the form of stars.

Despite their rarity, dust particles make interstellar space a relatively dirty place. Earth's atmosphere, by comparison, is about a million times cleaner. Our air is tainted by only one dust particle for about every billion billion (1018) atoms of atmospheric gas. If we could compress a typical parcel of interstellar space to equal the density of air on Earth, this parcel would contain enough dust to make a fog so thick that we would be unable to see our hand held at arm's length in front of us.


The composition of interstellar gas is reasonably well understood. Spectroscopic studies of interstellar absorption lines provide astronomers with comprehensive information on its elemental abundances. Most of it-- about 90 percent of all particles--is atomic and molecular hydrogen; some 9 percent is helium, and the remaining 1 percent consists of heavier elements. The abundances of several of the heavy elements, such as carbon, oxygen, silicon, magnesium, and iron, are much lower in interstellar gas than in our solar system or in stars. The most likely explanation for this finding is that substantial quantities of these elements have been used to form the interstellar dust, taking them out of the gas and locking them up in a form that is much harder to observe.

In contrast to interstellar gas, the composition of interstellar dust is currently not very well known. We have some infrared evidence for silicates, graphite, and iron--the same elements that are underabundant in the gas--lending support to the theory that interstellar dust forms out of interstellar gas. The dust probably also contains some "dirty ice," a frozen mixture of ordinary water ice contaminated with trace amounts of ammonia, methane, and other chemical compounds. This composition is quite similar to that of cometary nuclei in our own solar system (see Section 14.2).


Curiously, astronomers know the shape of interstellar dust particles better than their composition. Although the minute atoms in the interstellar gas are basically spherical, the dust particles are not. Individual dust grains are apparently elongated or rodlike, as shown in Figure 18.3. We can infer this because the light emitted by stars is dimmed and partially polarized, or aligned, by the dust.

Figure 18.3 A diagram of a typical interstellar dust particle. The average size of such particles is only 1/10,000 of a millimeter, yet space contains enough of them to obscure our view in certain directions.

Recall from Chapter 3 that light consists of electromagnetic waves composed of vibrating electric and magnetic fields. Normally, these waves are randomly oriented, and we say the radiation is unpolarized. Stars emit unpolarized radiation from their photospheres. Under some circumstances, however, the electric fields can become aligned--all vibrating in the same plane as the radiation moves through space. We then say the radiation is polarized. Polarization of starlight does not occur by chance. If the light detected by our telescope is polarized, it is because some interstellar matter lies between the emitting object and Earth. The polarization of starlight, then, provides another way to study the interstellar medium.

On Earth we can produce polarized light by passing unpolarized light through a Polaroid filter, which has specially aligned elongated molecules that allow the passage of only those waves with their electric fields oriented in some specific direction (see Figure 18.4a). Other waves are absorbed and do not pass through the filter. The alignment of the molecules determines which waves will be transmitted. In interstellar space, dust grains can act like the molecules in the Polaroid filter. If the starlight is polarized, astronomers can conclude that the interstellar dust particles must have an elongated shape (by analogy with the elongated molecules of the Polaroid filter) and that these molecules are aligned, as shown in Figure 18.4(b). Only then can the dust preferentially absorb certain waves, leaving the remainder (the ones we observe) polarized.

Figure 18.4 (a) Unpolarized light waves have randomly oriented electric fields. When the light passes through a Polaroid filter, only waves with their electric fields oriented in a specific direction are transmitted, and the resulting light is polarized. (b) Aligned dust particles in interstellar space polarize radiation in a similar manner. Observations of the degree of polarization allow astronomers to infer the size, shape, and orientation of the particles.

The alignment of the interstellar dust is the subject of intense research among astronomers. The current view, accepted by most, holds that the dust particles are affected by a weak interstellar magnetic field, perhaps a million times weaker than Earth's field. Each dust particle responds to the field in much the same way that small iron filings are aligned by an ordinary bar magnet. Measurements of the blockage and polarization of starlight thus yield information about the size and shape of interstellar dust particles, as well as about magnetic fields in interstellar space.

Now that we have a general idea of the basic contents and properties of interstellar space, let us examine some typical regions in more detail. We'll note especially how astronomers have unraveled the nature of the matter contained within them.

Back Figure 18.5 is a mosaic of photographs showing a region of space where stars, gas, and dust seem to congregate. The bright areas are made up of myriad unresolved stars. The dark areas are vast pockets of dust, blocking from our view what would otherwise be a rather smooth distribution of bright starlight. From our vantage point within the solar system, this assemblage of stellar and interstellar matter follows a bright band extending across the sky. On a clear night, this band of patchy light is visible to the naked eye as the Milky Way. In Chapter 23 we will come to recognize this band as the flattened disk, or plane, of our own Galaxy.

Figure 18.5 A mosaic of the plane of the Milky Way Galaxy. Photographed almost from horizon to horizon, and thus extending over nearly 180°, this band contains high concentrations of stars as well as interstellar gas and dust. The field of view is several times wider than that of Figure 18.1, whose outline is superimposed on this image.

Figure 18.6 shows a 12°-wide swath of the galactic plane in the general direction of the constellation Sagittarius, as photographed from Earth. The view is rather mottled, with a patchy distribution of stars and interstellar debris. In addition, several large fuzzy patches of light are clearly visible. These fuzzy objects, labeled M8, M16, M17, and M20, correspond to the 8th, 16th, 17th, and 20th objects in a catalog compiled by Charles Messier, an eighteenth-century French astronomer.* The stars, the fuzzy objects, and the dark obscuring matter are all concentrated around and along the galactic plane. Indeed, this plane is the site of greatest concentration of almost all astronomical objects within our Galaxy.

* Messier was actually more concerned with making a list of celestial objects that might possibly be confused with comets, his main astronomical interest. However, the catalog of 109 "Messier objects" is now regarded as a much more important contribution to astronomy than any comets Messier discovered.

Historically, astronomers used the term nebula to refer to any "fuzzy" patch (bright or dark) on the sky--any region of space that was clearly distinguishable through a telescope but not sharply defined, unlike a star or a planet. We now know that many (though not all) nebulae are clouds of interstellar dust and gas. If they happen to obscure stars lying behind them, we see a dark patch. If something within the cloud--a group of hot young stars, for example--causes it to glow, the nebula appears bright instead. The four fuzzy objects labeled in Figure 18.6 are emission nebulae--glowing clouds of hot interstellar gas. The method of spectroscopic parallax applied to stars visible within the nebulae indicates that their distances from Earth range from 900 pc (M20) to 1800 pc (M16). Thus, all four are near the limit of visibility for any object embedded in the dusty galactic plane. M16, at the top left, is approximately 1000 pc from M20, near the bottom.

We can gain a better appreciation of these nebulae by examining progressively smaller fields of view. Figure 18.7 is an enlargement of the region near the bottom of Figure 18.6. M20 is at the top, and M8 is at the bottom, only a few degrees away. Figure 18.8 is yet another enlargement of the top of Figure 18.7. This is a close-up of M20 and of its immediate environment--a real jewel of the night. The total area displayed measures some 10 pc across.

Figure 18.6 (Top left) A photograph of a small portion (about 12° across) of the galactic plane shown in Figure 18.1, displaying higher-resolution evidence for stars, gas, and dust as well as several distinct fuzzy patches of light, known as emission nebulae. The plane of the Milky Way Galaxy is marked with a dashed line.

Figure 18.7 (Bottom left) An enlargement of the bottom of Figure 18.6, showing M20 (top) and M8 (bottom) more clearly.

Figure 18.8 (Bottom right) Further enlargements of the top of Figure 18.7, showing only M20 and its interstellar environment. The nebula itself (in red) is about 4 pc in diameter. It is often called the Trifid Nebula because of the dust lanes that trisect its midsection (insert). The blue region is unrelated to the red emission nebula and is caused by starlight reflected from intervening dust particles. It is called a reflection nebula.

Emission nebulae such as M8 and M20 are among the most spectacular objects in the entire universe. Yet they appear only as small, undistinguished patches of light when viewed in the larger context of the entire galactic plane, as in Figure 18.5. Perspective is crucial in astronomy.

These nebulae are regions of glowing, ionized gas. At or near the center of each is at least one newly formed hot O- or B-type star producing copious amounts of ultraviolet light. As they travel outward from the star, ultraviolet photons collide with the surrounding gas, ionizing much of it. As electrons recombine with nuclei, they emit visible radiation, causing the gas to glow. Figure 18.9 completes our set of enlargements of the four nebulae visible in Figure 18.6. Notice in all cases the predominant red coloration of the emitted radiation and the hot bright stars embedded within the glowing nebular gas.

The reddish hue of these four nebulae--and in fact of any emission nebula--results from hydrogen atoms emitting light in the red part of the visible spectrum. Specifically, it is caused by the emission of radiation at 656.3 nm (6563 Å)--the H line that we encountered in Chapter 4. Other elements in the nebula also emit radiation as their electrons recombine, but because hydrogen is so plentiful, its emission overwhelms that from all other atoms. Tinged here and there with other colors, red usually dominates in emission nebulae.

Scientists often refer to the ionization state of an atom by attaching a roman numeral to its chemical symbol--I for the neutral atom, II for a singly ionized atom (missing one electron), III for a doubly ionized atom (missing two electrons), and so on. Because emission nebulae are composed mainly of ionized hydrogen, they are often referred to as HII regions. Regions of space containing primarily neutral (atomic) hydrogen are known as HI regions.

Woven through the glowing nebular gas, and plainly visible in Figures 18.7­18.9, are lanes of dark obscuring dust. Recent studies have demonstrated that these dust lanes are part of the nebulae and are not just unrelated dust clouds that happen to lie along our line of sight.

Figure 18.9 Enlargements of selected portions of Figure 18.6, showing (a) M16, the Eagle Nebula; (b) M17, the Omega Nebula and (c) M8, the Lagoon Nebula. Notice the irregular shape of the emitting regions, the characteristic red color of the light, the bright stars within the gas, and the patches of obscuring dust. Like the regions shown in Figures 18.7 and 18.8, these emission nebulae are the sites of recent star formation. Image (d) is a close-up of the core of M8, a region known as the Hourglass. (This photo was taken using a different filter--the detail that it reveals would not be visible in H light.)

The Eagle Nebula

M16, The Eagle Nebula

Gaseous Pillars of Star Birth

Close-up of Stellar "EGGs"


Most of the photons emitted by the recombination of electrons with atomic nuclei escape from the nebula. Unlike the ultraviolet photons originally emitted by the embedded stars, they do not carry enough energy to ionize the nebular gas, and they pass through the nebula relatively unhindered. Some eventually reach the Earth, where we can detect them. Only through these lower-energy photons do we learn anything about emission nebulae.

Just as studies of the spectra of ordinary stars contain a wealth of information about stellar atmospheres, nebular spectra tell us a great deal about ionized interstellar gas. Because at least one hot star resides near the center of the nebula, we might think that the combined spectrum of the star and the nebula would be hopelessly confused. In fact, it is not. We can easily distinguish nebular spectra from stellar spectra because the physical conditions in stars and nebulae differ so greatly. In particular, emission nebulae are made of hot thin gas that, as we saw in Chapter 4, yields detectable emission lines. When our spectroscope is trained on a star, we see a familiar stellar spectrum, consisting of a black-body-like continuous spectrum and absorption lines, together with superimposed emission lines from the nebular gas. When no star appears in the field of view, only the emission lines are seen.

Figure 18.10(b) is a typical nebular emission spectrum spanning part of the visible and near-ultraviolet wavelength interval. Numerous emission lines can be seen, and information on the nebula shown in Figure 18.10a can be extracted from all of them. The results of analyses of many nebular spectra show abundances close to those derived from observations of the Sun and other stars: Hydrogen is 90 percent abundant by number, followed by helium at about 9 percent; the heavier elements together make up the remaining 1 percent.

Figure 18.10 (a) The Butterfly Nebula, a glowing patch of gas a few parsecs across. (b) Its emission spectrum, showing light intensity as a function of frequency over the entire visible portion of the electromagnetic spectrum from red to deep violet.

Unlike stars, nebulae are large enough for their actual sizes to be measurable by simple geometry. Coupling this size information with estimates of the amount of matter along our line of sight (as revealed by the nebula's emission of light), we can find the nebula's density. Generally, emission nebulae have only a few hundred particles, mostly protons and electrons, in each cubic centimeter--a density some 1022 times lower than that of a typical planet. Spectral-line widths imply that the gas atoms and ions have temperatures around 8000 K. Table 18-1 lists some vital statistics for each of the nebulae shown in Figure 18.6.


When astronomers first studied the spectra of emission nebulae, they found many lines that did not correspond to anything observed in terrestrial laboratories. For a time, this prompted speculation that the nebulae contained elements unknown on Earth. Some scientists went so far as to invent the term "nebulium" for a new element, much as the name helium came about when that element was first discovered in the Sun (recall also "coronium" from Chapter 16). With a fuller understanding of the workings of the atom, astronomers realized that these lines did in fact result from electron transitions within the atoms of familiar elements. However, these transitions occurred under unfamiliar conditions not reproducible in laboratories.

For example, in addition to the dominant red coloration just discussed, many nebulae also emit light with a characteristic green color (see Figure 18.11). The greenish tint of portions of this nebula greatly puzzled astronomers in the early twentieth century and defied explanation in terms of the spectral lines known at the time.

Figure 18.11 (a) The Orion Nebula (M42) lies some 450 pc from Earth. It is visible to the naked eye as the fuzzy middle "star" of Orion's sword. (b) Like all emission nebulae, it consists of hot, glowing gas powered by a group of bright stars in the center. In addition to the red H emission, parts of the nebula show a slight greenish tint, caused by a so-called forbidden transition in ionized oxygen. (c) A high-resolution, approximately true-color image shows rich detail in a region about a light year across. Structural details are visible down to a level of 0.1 arc second, or 6 light hours--a scale comparable to our solar system.

A High-Resolution Mosaic of the Orion Nebula

Orion Nebular Mosaic

Orion Nebular Zoom

Slow Pan Across Orion Nebula

Astronomers now understand that the color is caused by a particular electron transition in doubly ionized oxygen. The structure of oxygen is such that an ion in the higher-energy state for this transition tends to remain there for a very long time--many hours, in fact--before dropping back to the lower state and emitting a photon. Only if the ion is left undisturbed during this time, and not kicked into another energy state, will the transition actually occur and the photon be emitted.

In a terrestrial experiment, no atom or ion is left undisturbed for long. Even in a "low-density" laboratory gas, there are many trillions of particles per cubic meter, and each particle experiences millions of collisions every second. The result is that an ion in the particular energy state that produces the peculiar green line in the nebular spectrum never has time to emit its photon--collisions kick it into some other state long before that occurs. For this reason, the line is usually called forbidden, even though it violates no laws of physics. It simply occurs on Earth with such low probability that it is never seen.

In an emission nebula, the density is so low that collisions between particles are very rare. There is plenty of time for the excited ion to emit its photon, and the forbidden line is produced. Numerous forbidden lines are known in nebular spectra. They remind us once again that the environment in the interstellar medium is very different from conditions on Earth and warn us of the problems of extending our terrestrial experience to the study of interstellar space.

Back Emission nebulae are only one small component of interstellar space. Most of space--in fact, more than 99 percent of it--is devoid of nebular regions, and contains no stars. It is simply dark. Look again at Figure 18.5, or just ponder the evening sky. The dark regions are by far the most representative regions of interstellar space. The gas density there is only about a million atoms per cubic meter--100 times lower than the density in an emission nebula. The temperature of this diffuse gas is typically around 100 K.

Within these dark voids between the nebulae and the stars lurks another type of astronomical object, the dark dust cloud. Such clouds are cooler than their surroundings and thousands or even millions of times denser. These clouds bear little resemblance to terrestrial clouds. Most of them are bigger than our solar system, and some are many parsecs across. (Yet even so, they make up no more than a few percent of the entire volume of interstellar space.) Despite their name, these clouds are actually made up primarily of gas, just like the rest of the interstellar medium. As just discussed, however, their absorption of starlight is due almost entirely to the dust they contain.


Figure 18.12 is an optical photograph of a typical interstellar dust cloud. Pockets of intense blackness mark regions where the dust and gas are especially concentrated and the light from background stars is completely obscured. This cloud takes its name from a nearby star, Rho Ophiuchi, and resides at the relatively nearby distance of about 300 pc. Measuring several parsecs across, this cloud is only a tiny part of the grand mosaic shown in Figure 18.5. Note especially the long "streamers" of (relatively) dense dust and gas. This cloud clearly is not spherical. Indeed, most interstellar clouds are very irregularly shaped. The bright patches within the dark region are emission nebulae in the foreground. Some of them are part of the cloud itself, where newly formed stars near the surface have created a "hot spot" in the cold, dark gas. Others have no connection to the cloud and just happen to lie along our line of sight.

Figure 18.12 Photograph of a typical interstellar dark dust cloud. This cloud, known as Rho Ophiuchi, is several parsecs across. It is "visible" only because it blocks the light coming from stars lying behind it. The approximate outline of the cloud is indicated by the dashed line.

These dark and dusty interstellar clouds are sprinkled throughout our Galaxy. We can study them at optical wavelengths only if they happen to block the light emitted by more distant stars or nebulae. The dark outline of Rho Ophiuchi and the dust lanes visible in Figures 18.8 and 18.9 are good examples of this obscuration. The dust is apparent only because it blocks the light coming from behind it. Figure 18.13 shows another striking example of a dark cloud--the Horsehead Nebula in Orion. This curiously shaped finger of gas and dust projects out from the much larger dark cloud in the bottom half of the image and stands out clearly against the red glow of a background emission nebula.

Figure 18.13 The Horsehead Nebula in Orion is a striking example of a dark dust cloud, silhouetted against the bright background of an emission nebula. The "neck" of the horse is about 0.25 pc across. The nebular region is roughly 1500 pc from the Earth.


Astronomers first became aware of the true extent of dark interstellar clouds in the 1930s as they studied the optical spectra of distant stars. In addition to the wide absorption lines normally formed in stars' lower atmospheres, much narrower absorption lines were also detected. Recall that the narrower the line, the cooler the temperature of the object absorbing the radiation. Figure 18.14(a) illustrates how light from a star may pass through several interstellar clouds on its way to Earth. These clouds need not be close to the star, and indeed they usually are not. Each absorbs some of the stellar radiation in a manner that depends on its own temperature, density, and elemental abundance. Figure 18.14(b) depicts part of a typical spectrum produced in this way.

Figure 18.14 (a) A simplified diagram of some interstellar clouds between a hot star and Earth. Optical observations might show an absorption spectrum like that traced in (b). The wide, intense lines are formed in the star's hot atmosphere; narrower, weaker lines arise from the cold interstellar clouds. The smaller the cloud, the weaker the lines. The redshifts or blueshifts of the narrow absorption lines provide information on cloud velocities. The widths of all the spectral lines depicted here are greatly exaggerated for the sake of clarity.

The narrow absorption lines contain information about dark interstellar clouds, just as stellar absorption lines reveal the properties of stars and nebular emission lines tell us about conditions in hot nebulae. By studying these lines, astronomers can probe the cold depths of interstellar space. In most cases, the elemental abundances detected in interstellar clouds mirror those found in other astronomical objects--which is perhaps not surprising, because interstellar clouds are the regions that spawn nebulae and stars. It appears that most of the matter in the Galaxy has become fairly well mixed by repeated processing in and out of stars, nebulae, and clouds. Most objects in the Galaxy have fairly similar composition.

Dark dust clouds are much cooler than the thousands of kelvins that characterize emission nebulae and stellar atmospheres. Their temperatures, as determined from their absorption line widths, are usually less than about 100 K, and values as low as 10 or 20 K are common. None of this interstellar gas can be ionized. It consists of neutral matter--atoms and molecules. This is why the clouds are invisible--they are just too cold to emit any visible light. They do, however, emit strongly at longer wavelengths. Compare the optical photograph of Rho Ophiuchi in Figure 18.12 with the infrared image of the same dark dust cloud in Figure 18.15, captured by sensitive detectors aboard the Infrared Astronomy Satellite.

Figure 18.15 A filled contour map of infrared radiation detected from the dark interstellar cloud Rho Ophiuchi. The infrared radiation, and therefore the dust that emits this radiation, displays a structure similar to the cloud's visual image (Figure 18.12). The very bright source of infrared radiation near the top of the cloud comes from a hot emission nebula, which can also be seen in the optical image. (The scale of the image is 3pc on a side; the black diagonal streak at right is an instrumental artifact.)

Along any given line of sight, cloud densities can range from 107 atoms/m3 to more than 1012 atoms/m3. These latter clouds are generally called dense interstellar clouds by researchers, but even these densest interstellar regions are about as tenuous as the best laboratory vacuum. Still, it is because their density is larger than the average value of 106 atoms/m3 that we can distinguish clouds from the surrounding expanse of interstellar space.

Back A basic problem with the optical technique just described is that we can examine interstellar clouds only along the line of sight to a distant star. To form an absorption line, there has to be a background source of radiation to absorb. The need to see stars through clouds also restricts this approach to relatively local regions, within a few thousand parsecs of Earth. Beyond that distance, stars are completely obscured, and no optical observations are possible. Only the denser, dustier clouds emit enough infrared radiation for astronomers to study them in that part of the spectrum.

To probe interstellar space more thoroughly, we need a more general, more versatile observational method--one that does not rely on conveniently located stars and nebulae. In short, we need a way to detect cold, neutral interstellar matter anywhere in space through its own radiation. This may sound impossible, but such an observational technique does in fact exist. The method relies on low-energy radio emissions produced by the interstellar gas itself.

Recall that a hydrogen atom has one electron orbiting a single-proton nucleus. Besides the electron's orbital motion around the central proton, electrons also have some rotational motion--that is, spin--about their own axis. The proton also spins. This model parallels a planetary system, in which, in addition to the orbital motion of a planet about a central star, both the planet (electron) and the star (proton) rotate about their axis.

The laws of physics dictate that there are exactly two possible spin configurations for a hydrogen atom in its ground state. The electron and proton can rotate in the same direction, with their spin axes parallel, or they can rotate with their axes antiparallel (that is, parallel, but oppositely oriented). Figure 18.16 shows these two configurations. The antiparallel configuration has slightly less energy than the parallel state.

Figure 18.16 Diagram of a ground-level hydrogen atom changing from a higher-energy state (electron and proton spins are parallel) to a lower-energy state (spins are antiparallel). The emitted photon carries away an energy equal to the energy difference between the two spin states.

All matter in the universe tends to achieve its lowest possible energy state, and interstellar gas is no exception. Even slightly excited hydrogen atoms eventually end up in their least energetic state. After a certain amount of time, hydrogen atoms in the higher-energy state relax back to the lower level. That is, the electron suddenly and spontaneously flips its spin, which then becomes opposite to that of the proton. As with any such change, the transition from a high-energy state to a low-energy state releases a photon with energy equal to the energy difference between the two levels.

The energy difference between the two states is very small, so the energy of the emitted photon is very low. Consequently, the wavelength of the radiation is rather long--in fact, about 21 cm, roughly the width of a textbook. That wavelength lies in the radio portion of the electromagnetic spectrum. Researchers refer to the spectral line that results from this hydrogen-spin-flip process as the 21-centimeter line. Figure 18.17 shows typical spectral profiles of 21-cm radio signals observed toward several different regions of space. These tracings are the characteristic signatures of cold, atomic hydrogen in our Galaxy. Needing no visible starlight to help calibrate their signals, radio astronomers can observe any interstellar region that contains enough hydrogen gas to produce a detectable signal. Even the low-density regions between the dark clouds can be studied.

Figure 18.17 Typical 21-cm radio spectral lines observed toward several different regions of interstellar space. The peaks do not all occur at a wavelength of exactly 21 cm, corresponding to a frequency of 1.4 GHz (gigahertz), because the gas in the Galaxy is moving with respect to Earth.

As can be seen in Figure 18.17, actual 21-cm lines are quite jagged and irregular, somewhat like nebular emission lines in appearance. These irregularities arise because there are usually numerous clumps of interstellar gas along any given line of sight. Each has its own density, temperature, radial velocity, and internal motion, so the intensity, width, and Doppler shift of the resultant 21-cm line vary from place to place. All these different lines are superimposed in the signal we eventually receive at Earth, and sophisticated computer analysis is generally required to disentangle them. The figures quoted earlier for the temperatures (100 K) and densities (106 atoms/m3) of the regions between the dark dust clouds are based on 21-cm measurements; observations of the dark clouds themselves yield densities and temperatures in good agreement with those obtained by optical spectroscopy.

All interstellar atomic hydrogen emits 21-cm radiation. But if all atoms eventually fall into their lowest-energy configuration, why isn't all the hydrogen in the Galaxy in the lower-energy state by now? Why do we see 21-cm radiation today? The answer is that the energy difference between the two states is comparable to the energy of a typical atom at a temperature of 100 K or so. As a result, atomic collisions in the interstellar medium are energetic enough to boost the electron up into the higher-energy configuration and so maintain comparable numbers of hydrogen atoms in either state. Any sample of interstellar hydrogen at any instant will contain many atoms in the upper level, and 21-cm radiation will always be emitted.

Of great importance, the wavelength of this characteristic radiation is much larger than the typical size of interstellar dust particles. Accordingly, this radio radiation reaches Earth completely unscattered by interstellar debris. The opportunity to observe interstellar space well beyond a few thousand parsecs, and in directions lacking background stars, makes 21-cm observations among the most important and useful in all of astronomy. We will see this technique used both in studies of our own Galaxy and in observations of truly distant astronomical objects.

Back In certain interstellar regions of cold (typically 20 K) neutral gas, densities can reach as high as 1012 particles/m3. Until the late 1970s, astronomers regarded these regions simply as abnormally dense interstellar clouds, but it is now recognized that they belong to an entirely new class of interstellar matter. The gas particles in these regions are not in atomic form at all; they are molecules. Because of the predominance of molecules in these dense interstellar regions, they are known as molecular clouds. Only within recent years have astronomers begun to appreciate the vastness of these clouds. They literally dwarf even the largest emission nebulae, which were previously thought to be the most massive residents of interstellar space.


As noted in Chapter 4, molecules can become collisionally or radiatively excited, much like atoms. Furthermore, again like atoms, molecules relax back to their ground states whenever the opportunity arises. In the process, they emit radiation. The energy states of molecules are much more complex than those of atoms, however. Molecules, like atoms, can undergo internal electron transitions, but unlike atoms, they can also rotate and vibrate. They do so in specific ways, obeying the laws of quantum physics.

Figure 18.18 illustrates a simple molecule that is rotating rapidly. After a length of time that depends on its internal makeup, the molecule relaxes back to a slower rotational rate (a state of lower energy). This change causes a photon to be emitted, carrying an energy equal to the energy difference between the two rotational states. The energy differences between rotational states are generally very small, so the emitted radiation is usually in the radio range.

Figure 18.18 As a molecule changes from a rapid rotation (left) to a slower rotation (right), a photon is emitted that can be detected with a radio telescope. Depicted here is the formaldehyde molecule, H2CO. The length of the curved arrows is proportional to the spin rate of the molecule.

We are fortunate that molecules emit radio radiation, because they are invariably found in the densest and dustiest parts of interstellar space. These are regions where the absorption of shorter-wavelength radiation is enough to prohibit the use of ultraviolet, optical, and most infrared techniques that might ordinarily detect changes in the energy states of the molecules. Only low-frequency radio radiation can escape, eventually to be detected on Earth.

Why are molecules found only in the densest and darkest of the interstellar clouds? One possible reason is that the dust serves to protect the fragile molecules from the normally harsh interstellar environment--the same absorption that prevents high-frequency radiation from getting out to our detectors also prevents it from getting in to destroy the molecules. Another possibility is that the dust acts as a catalyst that helps form the molecules. The grains provide both a place where atoms can stick and react and a means of dissipating any heat associated with the reaction, which might otherwise destroy the newly formed molecules. Probably the dust plays both roles. The close association between dust grains and molecules in dense interstellar clouds argues strongly in favor of this picture, but the details are still being debated.


By far the most common constituent of molecular clouds is molecular hydrogen (H2). Unfortunately, despite its prevalence, this molecule does not emit or absorb radio radiation. It emits only short-wavelength ultraviolet radiation, so it cannot easily be used as a probe of cloud structure. Nor are 21-cm observations helpful--they are sensitive only to atomic hydrogen, not to the molecular form of the gas. Theorists had expected H2 to abound in these dense, cold pockets of interstellar space, but proof of its existence was hard to obtain. Only when spacecraft measured the ultraviolet spectra of a few stars located near the edges of some dense clouds was the presence of molecular hydrogen confirmed.

With hydrogen effectively ruled out as a probe of molecular clouds, astronomers had to find other ways to study the dark interiors of these dusty regions. Fortunately, there are plenty to choose from. Spectral emissions caused by rotational changes of many different heavy molecules have been detected by radio telescopes. Molecules such as carbon monoxide (CO), hydrogen cyanide (HCN), ammonia (NH3), water (H2O), methyl alcohol (CH3OH), formaldehyde (H2CO), and about 60 others, some of them quite complex, are now known to exist in interstellar space.* Their abundances are very small--they are generally 1 million to 1 billion times less abundant than H2. But their parent molecular clouds are large and dense enough that photons emitted by many molecules accumulate to yield detectable signals.

* Some remarkably complex organic molecules have been found in the densest of the dark interstellar clouds, including formaldehyde (H2CO), ethyl alcohol (CH3CH2OH), methylamine (CH3NH2), and formic acid (H2CO2). Their presence has fueled speculation about the origins of life, both on Earth and in the interstellar medium--especially since the recent discovery by radio astronomers of evidence that glycine (NH2CH2COOH), one of the key amino acids that form the large protein molecules in living cells, may also be present in interstellar space.

These molecules are unimportant in terms of the overall properties of molecular clouds, but they play a vital role as tracers of a cloud's structure and physical properties. We believe that such molecules are produced by chemical reactions within molecular clouds--so when we observe high densities of formaldehyde, for example, we know that the regions under study also contain high densities of molecular hydrogen, dust, and other important constituents. Spectral studies of formaldehyde can thus provide physical information on the entire cloud. Moreover, the rotational properties of different molecules often make them suitable as probes of regions with different physical properties. Formaldehyde may provide the most useful information on one region, carbon monoxide on another, and water on yet another, depending on the densities and temperatures of the regions involved. These data equip astronomers with a sophisticated spectroscopic "toolbox" for studying the interstellar medium.

For example, Figure 18.19 shows some of the sites where formaldehyde molecules have been detected near M20. At practically every dark area sampled between M16 and M8, this molecule is present in surprisingly large abundance (although it is still far less common than H2). Analyses of spectral lines at many locations along the 12°-wide swath shown in Figure 18.6 indicate that the temperature and density are much the same in all the molecular clouds studied (50 K and 1011 molecules/m3, on average). Figure 18.20 shows a contour map of the distribution of formaldehyde molecules in the immediate vicinity of the M20 nebula. It was made by observing radio spectral lines of formaldehyde at various locations and then drawing contours connecting regions of similar abundance. Notice that the amount of formaldehyde (and, we assume, the amount of hydrogen) peaks in a dark region, well away from the visible nebula.

Figure 18.19 Spectra indicate that formaldehyde molecules exist around M20, as indicated by the arrows. The lines are most intense both in the dark dust lanes trisecting the nebula and in the dark regions beyond the nebula.

Figure 18.20 Contour map of the amount of formaldehyde near the M20 nebula, demonstrating how formaldehyde is especially abundant in the darkest interstellar regions. Other kinds of molecules have been found to be similarly distributed. The contour values increase from the outside to the inside, so the maximum density of formaldehyde lies just to the bottom right of the visible nebula. The green and red contours outline the intensity of the formaldehyde absorption lines at different rotational frequencies. The nebula itself is about 4 pc across.

Radio maps of interstellar gas and infrared maps of interstellar dust reveal that molecular clouds do not exist as distinct and separate objects in space. Rather, they make up huge molecular cloud complexes, some spanning as much as 50 pc across and containing enough gas to make millions of stars like our Sun. There are about 1000 such giant complexes known in our Galaxy.

The very existence of molecules has forced astronomers to rethink and reobserve interstellar space. In doing so, they have begun to realize that this active and interesting domain is far from the void suspected by theorists not so long ago. Regions of space recently thought to contain nothing more than galactic "garbage"--the cool, tenuous darkness among the stars--now play a critical role in our understanding of stars and the interstellar medium from which they are born.