CHAPTER 8

Earthlike Planets

BINARY STARS

A star may be Sunlike and yet still not be a suitable incubator for life. It may have properties, other than its mass and luminosity, that make it impossible for an Earthlike planet to circle it.

A star may be like the Sun in every apparent respect, for instance, and yet have as a companion not a planet or a group of planets, but another star. The presence of two stars in close association may conceivably ruin the chances for an Earthlike planet to circle either one.

The possibility of multiple stars did not dawn on astronomers until about two centuries ago. After all, our Sun is a star without stellar companions and that made it seem a natural condition. When the stars were recognized to be other suns, they, too, were assumed to be single. To be sure, there are stars that are close together in the sky. For instance, Mizar, the middle star in the handle of the Big Dipper, has a fainter star, Alcor, very near it. Such “double stars” were taken, however, to be single stars lying nearly in the same direction from the Earth but at radically different distances. In the case of Mizar and Alcor, this turned out to be true.

In the 1780s, William Herschel began to make a systematic study of double stars in the hope that the brighter (and presumably closer) one might move slightly and systematically with reference to the dimmer (and presumably more distant) one. This motion might reflect the motion of the Earth about the Sun and be the star’s “parallax.” From this, the star’s distance could be determined, something that had not yet been done.

Herschel did find motions among such stars, but never of the kind that would indicate the presence of a parallax. Instead, he found some double stars to be circling about a mutual center of gravity. These were true double stars, bound to each other gravitationally, and were called binary stars, from a Latin word meaning in pairs.

By 1802, Herschel was able to announce the existence of many such binary stars, and they are now known to be very common among the stars of the Universe. Among the bright and familiar stars, for instance, Sirius, Capella, Procyon, Castor, Spica, Antares, and Alpha Centauri are all binaries.

In fact, more than two stars might be held together gravitationally. Thus, the Alpha Centauri binary (which are referred to as Alpha Centauri A and Alpha Centauri B) have a very distant companion, Alpha Centauri C, some 1,600,000,000,000 kilometers (one trillion miles) from the center of gravity of the two other stars. A binary star system may also be gravitationally bound to another binary star system, the two pairs of stars circling a common center of gravity. Systems of five or even six stars are known.

In every case, though, where more than two stars are involved in a multiple system, the stars exist in relatively close pairs widely separated from companion singles or other binaries.

In other words, suppose that there were a planet about Star A, which is a member of a binary system. Star B might be close enough to have some important effect on the planet. It might add its own radiation to that of Star A in different amounts at different times. Or else its gravitational pull might introduce irregularities into the planet’s orbit that might not have existed otherwise.

On the other hand, if the A-B binary had, associated with it, a third star, or another binary, or both a star and a binary—all would be so far off that they would simply be stars in the sky without particular influence on the development of life on the planet.

From the standpoint of this book, therefore, let us talk only of binaries.

There is nothing puzzling about the existence of binaries.

When an initial nebula condenses to form a planetary system, one of the planets may, by the chance of the turbulence, attract enough mass to become a star itself. If, in the course of the development of our own Solar system, Jupiter had accumulated perhaps 65 times as much mass as it did, the loss of that mass to the Sun would not have been particularly significant. The Sun would still have very much the appearance it now has, while Jupiter would be a dim “red dwarf” star. The Sun would then be part of a binary system.

It is even quite possible that the original nebula might condense more or less equally about two centers to form stars of roughly equal mass, each smaller than our Sun, as in the case of the 61 Cygni binary system; or each roughly equal in size to our Sun, as in the case of the Alpha Centauri binary system; or each larger than the Sun, as in the Capella binary system.

The two stars might, if they are of different mass, have radically different histories. The more massive star may leave the main sequence, expand to a red giant, and then explode. Its remnants would then condense to a small, dense star, while the less massive companion star remains on the main sequence. Thus, Sirius has as a companion a white dwarf, a small, dense remnant of a star that once exploded. Procyon also has a white dwarf as a companion.

The total number of binaries in the Galaxy (and presumably in the Universe generally) is surprisingly large. Over the nearly two centuries since their discovery, the estimate of their frequency has steadily risen. At the moment, judging from the examples of those stars close enough to ourselves to be examined in detail, it would seem that anywhere from 50 to 70 percent of all stars are members of a binary system. In order to arrive at a particular figure, let us take an average and say that 60 percent of all stars and, therefore, of all Sunlike stars, too, are members of a binary system.

If we assume that any Sunlike star can form a binary with a star of any mass, then, keeping in mind the proportions of stars of various masses, we could venture a reasonable division of the 75 billion Sunlike stars in the Galaxy as follows:

Ought we now to eliminate the 45 billion Sunlike stars involved in binary systems as unfit incubators for life?

Certainly, it would seem that we can omit the 2 billion Sunlike stars that form binaries with giant stars. In their case, long before the Sunlike star has reached an age where intelligence might develop on some planet circling it, the companion star would explode as a supernova. The heat and radiation of a nearby supernova is quite likely to destroy any life on the planet that already existed.

What about the remaining 43 billion Sunlike stars forming a part of binaries?

In the first place, can a binary system possess planets at all?

We might argue that if a nebula condenses into two stars, the two will be twice as effective in picking up debris as one would be. Any planetary material that might escape one would be picked up by the other. In the end, therefore, there would be two stars and no planets.

That this is not necessarily so is demonstrated by the star 61 Cygni, the first whose distance from Earth was determined, in 1838, and that is now known to be 11.1 light-years from us.

61 Cygni, as I have said earlier, is a binary star. The two component stars, 61 Cygni A and 61 Cygni B, are separated by 29 seconds of arc as viewed from Earth (a separation about 1/64 the width of the full Moon).

Each of the component stars is smaller than the Sun, but each is large enough to be Sunlike. 61 Cygni A has about 0.6 times the mass of the Sun, and 61 Cygni B about 0.5 times the mass. The former has a diameter of about 950,000 kilometers (600,000 miles) and the latter a diameter of about 900,000 kilometers (560,000 miles). They are separated by an average distance of about 12,400,000,000 kilometers (7,700,000,000 miles), or a little over twice the average distance between the Sun and Pluto, and they circle each other about their center of gravity once in 720 years.

If we imagined the planet Earth circling one of the 61 Cygni stars at the same distance it now circles the Sun, the other 61 Cygni star would appear in the night sky at various times as a bright, starlike object, showing no visible disc, delivering no significant amount of radiation, and producing no significantly interfering gravitational effect.

Indeed, we might easily imagine each 61 Cygni star as possessing a planetary system nearly as extensive as the Sun’s, each without interference from the other.^* In this particular case, we need not resort entirely to speculation. The very first planetary object about another star for which some evidence was obtained involved 61 Cygni. From the manner in which the separation of the two stars changed in a wobbly manner as they circled each other, the presence of a third body, 61 Cygni C, was deduced. From the extent of the wobble, it was thought to be a large planet some eight times the mass of Jupiter.

Soviet astronomers at the Pulkovo Observatory near Leningrad have studied the orbits of the 61 Cygni stars with care, have measured the irregularities of the wobble itself, and have suggested, in 1977, that three planets are involved. They feel that 61 Cygni A has two large planets, one with 6 times the mass of Jupiter and one with 12 times the mass, while 61 Cygni B has one large planet with 7 times the mass of Jupiter.

These are very borderline observations. The tiny changes in the motion of the 61 Cygni stars can just barely be made out, and the chance that insignificant errors of measurement or interpretation have produced them is all too likely.

For what it’s worth, however, and until something better comes along, it implies that both stars of a binary system (both stars being Sunlike stars) have planets—large planets at least. If large planets exist, however, it doesn’t take much of a strain to suppose the existence of a large collection of smaller planets, satellites, asteroids, and comets—all too small to leave detectable marks on the wobble.

Of course, some binary systems are separated by smaller distances than the 61 Cygni stars.

Consider the two stars of the Alpha Centauri binary system. Alpha Centauri A has a mass 1.08 times that of the Sun, and Alpha Centauri B a mass 0.87 times that of the Sun. The two stars are separated by an average distance of 3,500,000,000 kilometers (2,200,000,000 miles). They revolve about the center of gravity in quite elliptical orbits, however, and are much closer to each other at some times than at others. The maximum distance between the two stars is 5,300,000,000 kilometers (3,400,000,000 miles) and the minimum distance between the two is 1,700,000,000 kilometers (1,050,000,000 miles).

Suppose we imagined Alpha Centauri B circling our Sun exactly as it, in fact, circles Alpha Centauri A. If we plotted Alpha Centauri B’s orbit relative to the Sun, it would follow an elliptical path that would carry it well beyond the orbit of Neptune at its farthest recession from the Sun, and nearly as close as the orbit of Saturn as its nearest approach.

Under such circumstances, neither star could have a very extensive planetary system of the sort the Sun has now. Planets at the distance of Jupiter or the other giants, circling either star, would be interfered with by the gravitational influence of the other star and would have unstable orbits.

On the other hand, an inner planetary system might still exist. If Alpha Centauri B were circling our Sun as it circles Alpha Centauri A, we on Earth could scarcely tell the difference with our eyes closed. Alpha Centauri B would be a bright, starlike object in the sky, which at its closest approach would be 5,000 times brighter than our full Moon and 1/100 as bright as our Sun. It would add anywhere from 0.1 percent to 1 percent to the heat we receive from the Sun, depending on what part of its orbit it was in, and we could live with that. Nor would its gravitational influence affect Earth’s orbit in any significant way.

For that matter, Alpha Centauri B could have an inner planetary system, too. A planet circling in its ecosphere (which would of course be closer to itself than the ecosphere is to either Alpha Centauri A or the Sun) would not be seriously interfered with by its somewhat larger companion.

As in the case of the 61 Cygni system, both Alpha Centauri A and Alpha Centauri B would have what we might call a “useful ecosphere,” one in which an Earthlike planet could orbit without serious interference from the companion in terms of either radiation or gravitation.

Robert S. Harrington of the U.S. Naval Observatory in 1978 reported the results of high-speed computer studies of orbits about binary stars.

If a Sunlike star is part of a binary system, and if the separation between the two stars is at least 3.5 times the distance of the ecosphere from the Sunlike star, then it is a useful ecosphere. In the case of our own Solar system, it would mean that the Sun could have a companion at a distance equal to that of the planet Jupiter, without interfering with Earth gravitationally. If the companion were somewhat less luminous than Alpha Centauri B, it would not interfere with Earth significantly as far as radiation was concerned.

There are binary systems with stars even closer together than those of the Alpha Centauri system. The two stars of the Capella binary system are separated by a distance of only 84 million kilometers (52 million miles) or less than the distance of Venus from the Sun.

Neither star in such a binary could have a planetary system in the Sun’s sense. Planetary orbits about one of the stars would be interfered with gravitationally by the other and the orbit would not be stable.

If a planet were far enough away, however, it would not circle either one star or the other but would circle, instead, about the center of gravity of the two stars. Such a planet would treat the two stars gravitationally as a single dumbbell-shaped object.

Harrington calculates that a planet whose distance from the center of gravity of the binary system was equal to at least 3.5 times the distance of separation between the two stars would have a stable orbit. In the case of the Capella system, a planet, to have a stable orbit, would have to be at least 300 million kilometers (185 million miles) from the center of gravity.

In a close binary system, where the two stars are of the proper total luminosity, such an outer orbit might well be within the ecosphere of the two stars taken together. This is another way in which a binary might have a useful ecosphere.

There are pairs of stars that circle each other so closely that our best telescopes cannot make them out as separate stars. Their existence as pairs is given away by the spectroscope, when the dark lines of the spectrum sometimes double, rejoin, double, rejoin, and so on, over and over.

The simplest explanation is to suppose that there are two stars very close together and circling each other, so that one is receding from us while the other is approaching us. In that case, one would produce a red shift, while the other was simultaneously producing a violet shift, and that is why the lines would appear to double. It is the same principle that causes the lines of a rotating star to broaden. The revolution of two stars is more rapid than the rotation of one star, so that in the latter case the broadening is carried on to the point of actual spreading apart into two lines.

The first such “spectroscopic binary” to be discovered was Mizar, and it was in 1889 that the American astronomer Edward Charles Pickering (1846–1919) detected the doubling of its spectral lines. Actually, the component stars of Mizar are separated by 164 million kilometers (102 million miles), which is a larger separation than that of the stars of the Capella system. The Mizar pair fail to be seen as a pair in the telescope because the system is so far away.

The component stars of some spectroscopic binaries are much closer to each other than that. They can be within a million kilometers of each other, almost touching, and making a complete circle about the center of gravity in a couple of hours.

If we could imagine the Sun replaced by two stars, each half as luminous as the Sun and separated by less than 42,700,000 kilometers (26,500,000 miles)—somewhat less than the distance between the Sun and Mercury—the Earth would remain stably in its orbit. Planets at the distance of Mercury and Venus could not, under those conditions, remain in stable orbit, but Earth could.

In such a case, of course, the sum of the mass of the two stars would be greater than that of the Sun, and Earth’s period of revolution would be considerably less than a year. In addition, with two separate stars at changing distances, Earth’s seasons would show more complicated variations, perhaps, than they now do. Neither of these two factors, however, need render Earth unsuitable for life.

Well, then, how many Sunlike stars in our Galaxy have useful ecospheres?

To begin with, we may fairly assume that all the Sunlike stars that are single have useful ecospheres, and that means 30 billion right there.

Of the binary systems we have eliminated all Sunlike stars that have as a companion a giant star (or a small, dense star that is the shrunken and condensed remnant of a giant star that exploded).

Of the 18 billion Sunlike stars that are in binary association with another Sunlike star, we might estimate conservatively that only one-third have useful ecospheres. That would mean 6 billion stars in this category. At a guess (nothing more than that) I would say there would be 4 billion binaries with two Sunlike stars, in which only the larger would have a useful ecosphere; and one million binaries of this kind in which both Sunlike stars would have a useful ecosphere.

Finally, what of the binaries in which a Sunlike star is teamed with a midget star? We had estimated there were 25 billion such binaries in the Galaxy altogether. A midget star is far less likely to interfere with a planetary system, either gravitationally or radiationally, than a larger star would. We might estimate, again conservatively that two-thirds of these Sunlike stars have useful ecospheres, and this would mean approximately 16 billion stars.

We now have our fourth figure:

STAR POPULATIONS

Yet we are not through. A Sunlike star may have a useful ecosphere and even so there may be no possibility of an Earthlike planet revolving within that ecosphere. As it happens, stars may differ in ways other than mass, luminosity, and state of association. They may also differ in chemical composition.

When the Universe first formed about 15 billion years ago, matter seems to have spread outward from an exploding central mass. To begin with, that matter consisted almost entirely of hydrogen, the simplest element, with a small admixture of a few percent of helium, the next simplest element. Virtually none of the still heavier elements existed.

This primordial matter, forming a Universe-sized mass of gas, split up into turbulent sections, each of galaxy size. Out of these protogalaxies, the stars of the various galaxies formed.

If we concentrate on any of the galaxy-sized masses of gas, the central regions were denser than the outer regions. The gas in the central regions split up into small, star-sized masses pretty evenly, each crowding the other so that no one star-sized mass had more chance than another to collect its share. The result was that very many stars were formed, all small and medium in size; virtually none of them giants. What’s more, nearly all the gas was collected by one star or another, so that the interstellar regions in a galactic center ended up almost gas free.

These stars, characteristic of the central regions of a galaxy, are called Population II stars.

For regions at moderate distance outside the center, there is not enough gas to form a steady, continuous packing of stars. The gas shreds into a couple of hundred smaller pockets of denseness, however, and out of each of them a tight group of some ten thousand to a million stars form. In this way, a “globular cluster” is formed. Globular clusters are arranged in a spherical shell about the galactic center, and are virtually dust free; the stars in such clusters are also Population II in nature.

The point to remember about Population II stars is that they were formed out of a gas that was largely hydrogen, with a little bit of helium, and virtually nothing else. The planetary systems that formed about such stars must be made up of planets that are also of that chemical structure. What planets do form about Population II stars would rather resemble Jupiter and Saturn in composition, but would lack the admixture of ices—water, ammonia, methane, and so on—that those planets possess.

There would be no small objects in the planetary systems, since small objects would not have enough gravitational pull to retain the hydrogen and helium which were alone available.

Nor would there be life, for to have life (as we know it) we need such elements as carbon, oxygen, nitrogen, and sulfur, which are not present in appreciable amounts in Population II planetary systems.

Of course, the heavier elements do form with time. As each Population II star burns over the course of the billions of years, heavy elements build up in its core through fusion reactions, including particularly those needed for life.

These heavier elements are, however, useless for the production of life as long as they remain at the core of stars.

Eventually a star leaves the main sequence, however, expands, and then collapses. If the star is a small one and not too much larger than our Sun, the process of collapse is not accompanied by an explosion, and a white dwarf is produced. In the process of collapse, however, up to one-fifth of the mass of the collapsing star is left behind as a cloud of gas surrounding the white dwarf. The result is what is called a planetary nebula. The expanding shell of gas slowly spreads through space until it becomes too rarefied to detect visually, and left behind is a bare white dwarf.

If a star is more massive than 1.4 times the mass of the Sun, it explodes as it collapses. The more massive the star, the more violent the explosion. Such a supernova explosion can eject up to nine-tenths of the mass of a star into space as swirls of gas.

The gas spreading into space, whether it started as the product of a planetary nebula or of a supernova, contains appreciable percentages of the more complicated elements. The process of supernoval explosion would, in particular, manufacture the really complex elements, which do not form in the center of stars that are quietly maturing on the main sequence. In the center of those stars, nothing past iron is produced, whereas in the comparatively brief episode of the supernova explosion, elements up to uranium and beyond are produced.

The Population II stars, however, are not very massive and, containing as they do a high percentage of hydrogen to begin with, they remain on the main sequence for a long time. Even in the 15 billion years that have elapsed since the big bang, almost all of those stars are still on the main sequence and the heavy elements remain tucked inside their cores.

From all this we might deduce that the centers of galaxies are quiet, uneventful places—and we would be wrong.

In 1963, quasars were discovered. These are starlike objects; indeed, when first discovered they were thought to be dim stars of our own Galaxy. They turned out, instead, to be located at distances of over a billion light-years, farther than any of the visible galaxies. To be visible at that distance, quasars had to be shining with the luminosity of 100 ordinary galaxies. Yet they are small objects, at most one or two light-years across, as compared with the diameters of many thousands of light-years that characterize ordinary galaxies.

The evidence now seems to favor the thought that quasars are bright galactic centers, surrounded, of course, by the outer structure of an ordinary galaxy. At the huge distance of the quasars, however, only the bright center is visible.

The question, then, is: what makes a galactic center blaze so brightly?

It would appear that the very centers of galaxies are quite commonly the sites of violent events. Some are visibly exploding; some give off vast streams of radio waves from sources on either side of the center as though an explosion has ejected material in opposite directions.

All galactic centers are bright; some are brighter than others. As we consider galaxies that are more and more distant, we reach a point where we see only the brightest of the bright galactic centers—the quasars.

What happens to the quiet Population II stars to initiate such violence?

If they were left to themselves, nothing; but they are not left to themselves. In the crowded precincts of the galactic centers, the stars are a million times as densely packed as in our own area of the galactic outskirts. The stars at the galactic center may be separated by average distances of only 70 billion kilometers (45 billion miles), only ten times the distance between the Sun and Pluto.

Under such packed conditions, collisions and near-collisions may not be very rare. Transfer and capture of mass may serve to build up stars of great mass that quickly explode with a force that leads to a veritable chain reaction of explosions and to the formation of “black holes.” These are the ultimate in star condensations (see my book, The Collapsing Universe).

A black hole is matter at its ultimate density, and has a gravitational field so intense at its surface that nothing can escape it, not even light.

If a black hole is formed under conditions in which matter of all kinds surrounds it (as in galactic centers), such matter is constantly spiraling into the black hole, releasing x-rays and other energetic radiation in the process. (This radiation is released before the matter actually enters the black hole, so that it can escape into outer space.) The black hole gains in mass and may eventually be large enough to swallow stars whole.

There is a strong radiation source at the very center of our own Galaxy, and it may well be that a black hole is present there, one that has a mass of 100 million stars. The giant galaxy M87 was reported in 1978 to have a black hole in its center in all likelihood, one that has a mass as high as that of 10 billion stars. It may even be that every galaxy and every globular cluster has a black hole at its core.

Such violent events at the centers of galaxies may produce the massive atoms of complex elements and spread them through space, but of what use would that be? Those violent events are the sites of emission of enormous quantities of energetic radiation, and for many light-years in every direction, life (as we know it) might for that reason be impossible.

The Population II regions are therefore, considering chemical constitution or energetic radiation, doubly unsuitable for life.

Suppose we pass on to the outskirts now, regions where the violence and radiation of the center does not reach.

Here, the primordial gas was relatively thin and was distributed irregularly. For that reason, stars were formed irregularly, and giant stars were routinely formed in numbers that could not possibly have existed in the center. (Of course, many medium and small stars were also formed.)

The stars in the outskirts of a galaxy, rich in giants and spread out irregularly over much vaster volumes of space than exist in the central regions, are referred to as Population I stars.^* What’s more, there were places in the outskirts where the gas was too thin to condense readily. To this day, therefore, the outer Population I regions of the galaxies are rich in clouds of gas and dust.

The original Population I stars were as entirely hydrogen-helium in constitution as were the Population II stars. There was this difference, however:

The giant stars that formed in the galactic outskirts didn’t remain on the main sequence long. A few hundred thousand years only, for the real monsters; a few million years for the mere titans; as much as a billion years for those that were simply giant.

And when they left the main sequence, expanded and finally collapsed, they exploded into supernovas of unimaginable violence. Vast volumes of gas, containing significant quantities of complex elements rolled out into space, adding themselves to the clouds of uncondensed gas that were already present.

Such explosions take place repeatedly in the outer regions of a galaxy, but so widely separated are the stars in those vast outer regions that supernovas do not seriously affect any stars other than (at most) their immediate neighbors.

As many as 500 million supernova explosions may have taken place in the outskirts of our own Galaxy since it came into being. The 500 million have enriched space enormously with complex elements, and have added to the density of the clouds of gas and dust that existed from the beginning. The outward force of the explosion may even have served as an initiation of swirls and compressions in nearby gas clouds that led to the formation of a new star, or whole groups of new stars.

New stars, forming out of gas clouds containing elements produced in an older star that had distributed those elements in its death throes, are called second-generation stars. Our Sun, which formed only 5 billion years ago, when the Galaxy was already 10 billion years old and after hundreds of millions of stars had already died, is a second-generation star.

The cloud out of which second-generation stars are formed contain the elements out of which ices, rocks, and metals are formed, and therefore can produce planetary systems similar to our own Solar system.

If we look for Sunlike stars that are capable of incubating life, therefore, we must eliminate Population II stars and even many of the Population I stars. We can only consider second-generation Population I stars.

Population II stars are confined to only a small portion of the total volume of a galaxy, to its compact central regions and to the almost as compact globular clusters. All the open vastness of the outer regions is the domain of Population I stars.

That is not, however, as impressive as it sounds. Some 80 percent of the stars of a galaxy are to be found in the compact central regions and in the globular clusters.

We might argue, too, that only half of the 20 percent of stars that are in the Population I regions are second-generation stars. That means that only 10 percent of all the Sunlike stars with effective ecospheres are second-generation Population I stars, and can conceivably have Earthlike planets revolving about them.

That gives us our fifth number:

THE ECOSPHERE

Even if a star is a perfect incubator, if it is the precise duplicate of our Sun in every respect, that is still not enough. What is needed is not only an incubator, but something to be incubated as well. In short, there must be a planet on which life can develop in the beneficent radiation of the star it circles.

To be sure, we have already decided that virtually every star has its planetary system, so that there are 5,200,000,000 second-generation, Population I, Sunlike stars in our Galaxy with planets—but where are those planets located?

A given star might be a perfect incubator, but some of its planets may be too close to it and therefore too hot to bear life, while others might be too far and therefore too cold to bear life. There might be no planet at all within the star’s ecosphere on which water could exist as a liquid.

What are the chances, then, that a given star has a planet, at least one, within its ecosphere?

In trying to make a judgment here, we are badly hampered by the fact that we know only one planetary system in detail—our own. What’s more, we have no way at all at present of possibly learning any appropriate details about any other planetary system. The few planets we may possibly have detected circling nearby stars are all the size of Jupiter or larger.

Such giant planets are the only ones we can possibly detect at the moment, and that only with great difficulty and considerable uncertainty. Whether there are any planets actually within the ecosphere of such stars, planets that lie closer to the star and that are small enough to be Earthlike, it is impossible to tell.

We are forced to fall back on the only thing we have, our own planetary system. It may conceivably be a very atypical, freakish structure that simply can’t be used as a guide, but we have no reason to think so. The temptation is to follow the principle of mediocrity and to suppose that the planetary system in which we find ourselves is a typical one and that it can be used as a guide.

There is some hope that this is not just prejudice on our part, or wishful thinking. The American astronomer Stephen H. Dole has checked this, as well as one can, by computer. Beginning with a cloud of dust and gas of the mass and density thought to have served as the origin of the Solar system, he set up the requirements for random motion, for coalescence on collision, for gravitational effects, and so on. The computer calculated the results.

The computer worked our different random happenings, and in every case a planetary system very much like ours resulted. There were from seven to fourteen planets, with small planets near the Sun, large planets farther out, and small planets again still farther out. In almost every case, there was a planet rather like the Earth in mass, at rather Earth’s distance from the Sun, and planets much like Jupiter in mass at much like Jupiter’s distance from the Sun, and so on.

In fact, if a diagram of the real Solar system is mixed in with the various computer simulations, it is not at all easy to separate the real from the simulated.

It is hard to say how much importance we can lend to such computer simulations, but for what they are worth, they do give a color of truth to the principle of mediocrity, at least in this respect.

If we now study our own planetary system on the assumption that it is typical, we can see that the planets move in nearly circular orbits that are widely spaced, and that the orbit of one does not overlap the orbit of the planet within or the one without.

This tends to make sense, since orbits that are too closely spaced would, in the long run, prove unstable. Between collisions and gravitational interactions, the worlds are bound to nudge themselves apart early in the history of the planetary system.

This means that it is completely unlikely that there will be very many worlds crammed into the ecosphere of a Sunlike star. The ecosphere is not likely to be wide enough for that. In fact, we might suspect intuitively that once the planets are done nudging themselves apart, not more than one planet is likely to find itself within the ecosphere; or two, if we find ourselves dealing with a double planet on the order of the Earth and the Moon.

How does this check with our own planetary system?

Here, for instance, Earth is clearly within the Sun’s ecosphere, or you and I would not exist to question the matter.

Even as late as a generation ago, the ecosphere would have seemed to be some 100 million kilometers (62 million miles) deep at least, since it was generally supposed that while Venus might be uncomfortably warm and Mars uncomfortably cool, both had environments not so extreme as to preclude the presence of life.

Not so. Venus has suffered a runaway greenhouse effect and is far too hot for life. Mars may be in a permanent ice age and be far too cold for life. The trigger leading in either direction may be a minor one.

If this is so, the Sun’s ecosphere may be shallower than we think. Indeed, in 1978, Michael Hart of NASA simulated Earth’s past history by computer and if his starting assumptions are correct, and his computer programming likewise, then it would seem that Earth, at one stage in its history, escaped a runaway greenhouse effect by a narrow margin and at another stage escaped a runaway ice age by a narrow margin. A little nearer the Sun or a little farther from it, and Earth would have fallen prey to one or the other. It may be, from Hart’s figures, that the Sun’s ecosphere is only 10 million kilometers (6,200,000 miles) thick and it is only a most fortunate coincidence that Earth happens to be in it.

Well, then, what can we say? If the ecosphere is quite wide (even if not wide enough to include either Venus or Mars), then from Dole’s computer simulation of planetary systems, a planet is virtually certain to form within it somewhere. The probability would be roughly 1.0.

On the other hand, if Hart’s computer simulation of Earth’s past history is accurate, then it is very likely that no planet at all will form within the ecosphere, and that all the planets near the star will be Venuslike or Marslike, and only on quite rare occasions Earthlike. The probability of a planet within the ecosphere would then be close to 0.0.

The results of computer simulation are still too recent and, perhaps, too crude to allow us to lean too certainly in either the optimistic or the pessimistic direction. It might be best to split the difference and to suppose that the probability of a planet within the ecosphere is close to 0.5, or 1 in 2.

That would give us our sixth number:

HABITABILITY

The mere fact that a planet is in the ecosphere does not mean that it is a suitable abode for life; that it is habitable, in other words.

For proof of that we need look no farther than our own Solar system. The Earth itself is the only planet in the Solar system that is clearly within the ecosphere of the star it circles. Our definition of the word planet, however, obscures the fact that there are two worlds in the ecosphere just the same.

The Moon, strictly speaking, is not a planet, because it circles the Earth (or rather the Earth-Moon center of gravity, which the Earth also circles), but it is a world. What’s more, it is a world that is just as firmly within the ecosphere as the Earth and yet the Moon is not a habitable world^.* The Moon clearly has too little mass to be habitable, since because of its small mass it cannot retain an atmosphere or liquid water. What, then, can we say about the masses of planets?

As I have said in the case of Population II stars where the only materials for planetary structure are hydrogen and helium, the only possible planets would seem to be giants with the mass of Uranus or more. Nothing less would possess the gravitational intensity that would make it possible to hold on to hydrogen and helium.

In the case of Population I stars, which are the only ones we are considering as suitable incubators for life, we have metals, rocks, and ices in addition to hydrogen and helium for uses as structural materials. Again here, only giant planets can make use of the hydro-gen and helium, and it is precisely because they can that they are giant planets.

On the other hand, where Population I stars are concerned, smaller worlds of all sizes can be built up of metals, rocks, and ices, since these will hold together through forces other than gravitational.

How large can these smaller worlds be?

Not very large, for even among Population I stars of the second generation, the quantity of materials other than hydrogen and helium is rather small, and cannot be used to build a large world. And if these stars could, they would gather hydrogen and helium and become giant worlds.

Dole’s computer simulations of planetary formation make it seem pretty clear that within the ecosphere of Sunlike stars those planets that are not giants are quite small.

How large and massive can a nongiant planet be?

If we exclude the four giant planets of the Solar system (and the Sun itself, of course), then the largest body in the Solar system is none other than the Earth itself.

Earth is, therefore, very likely to be near the top limit of mass for nongiant, nonhydrogen planets.

A planet somewhat larger than Earth, but not much larger, would, if all other factors were suitable, surely be habitable. The one unavoidable consequence of the greater mass would be a more intense gravitational field, which might manifest itself as a somewhat higher surface gravity. There is no reason to think that life could not adapt itself to a somewhat higher surface gravity.

After all, life on Earth evolved in the ocean where, thanks to buoyancy, the influence of gravity is minor. Living organisms invaded the dry land, where the influence of gravity is major, yet not only coped with it but even evolved ways of moving rapidly despite gravity. A somewhat greater surface gravity would surely not defeat life when it has shown such amazing adaptability on the one world where we can study it in detail.

Then, too, if a world is somewhat more massive than Earth, but also somewhat less dense, so that its surface is farther from the center than one would expect under Earthlike conditions, the surface gravity may be no higher than that of Earth, or even a bit lower.

We might reasonably conclude, then, that in the ecosphere where a star’s heat will be great enough to preclude the gathering of hydrogen and helium, planets will not form that are too massive for life.

Worlds that are not massive enough can certainly form, as for instance the Moon, but how massive is not massive enough?

To support life, a world must be massive enough to generate a sufficiently large gravitational field to hold a substantial atmosphere—not so much for the sake of the atmosphere, as because that alone would make it possible to have free liquid on the surface.

In the Solar system there are exactly four of the nongiant worlds with substantial atmospheres: Earth, Venus, Mars, and Titan.

Venus, with a mass 0.82 that of the Earth, has a considerably denser atmosphere than Earth (but is nonhabitable for other reasons). Mars, which has 0.11 times the mass of the Earth, has a very thin atmosphere; one that, while substantial, is clearly not sufficient to support anything but, just possibly, the simplest forms of life. Titan, which has a mass 0.02 that of Earth, has an atmosphere that may be somewhat more substantial than that of Mars, but which exists at all only because Titan is far beyond the outermost reach of the ecosphere.

Within the ecosphere, a world can maintain an adequate atmosphere if it is not as massive as Earth, but it should certainly be more massive than Mars. If, let us say, its mass were 0.4 times that of Earth, that might be sufficient.

In or near the Sun’s ecosphere, there are four worlds of considerable size: Earth, Venus, Mars, and the Moon. (There are also bodies of trifling size, such as the two satellites of Mars, and periodic entries of asteroids or comets, but these may all be ignored as not significant.) Of these four, Earth and Venus are higher in mass than the 0.4 mark, while Mars and the Moon are lower.

If we use the principle of mediocrity and consider this as a fair sample of the situation in the Universe as a whole, we could conclude that of all the worlds in or near appropriate ecospheres surrounding appropriate stars, only half have masses suitable for habitability.

If a world of the proper mass is present in the ecosphere, many of its characteristics would automatically be like those of the Earth. For instance, it would be too warm for substantial quantities of the icy materials to be in the solid state; and in liquid or gaseous state, the gravitational field of the world would not be intense enough to hold them. Therefore, a world of the proper mass in the ecosphere would be built up primarily of rock, or of rock and metal, as are all the worlds of the inner Solar system.

Water, as the icy material that melts and boils at the highest temperature, that is the most common, and that most readily combines with rocky substances, is on all three counts the most likely of the ices to be retained to some degree. Therefore, worlds of the proper mass in the ecosphere are very likely to have quantities of surface water in gaseous, liquid, and solid form. They would have oceans that would cover at least part of the surface.

In short, a world in the ecosphere that is of the proper mass would be “Earthlike” in character.

If one out of every two worlds in the ecosphere is Earthlike, we have our seventh figure:

Even an Earthlike planet, in terms of temperature and structure, might be nonhabitable for any of a variety of minor reasons. It could not very well support life if it were subjected, for instance, to great extremes of environmental conditions.

Suppose a planet had an average distance from the Sun that was right in the middle of the ecosphere, but suppose it also had a particularly eccentric orbit. At one end of its orbit it might swoop so far toward the Sun as to be well inside the inner border of the ecosphere, while on the other side it would recede so far from the Sun as to be well outside the outer border. Such a planet would have a short, unbelievably torrid summer that might briefly bring the oceans to a boil; and a long, unbelievably frigid winter, during which the oceans may begin to freeze.

We can imagine life might develop that could withstand such extremes, but it seems reasonable to suppose that the chances are it would not.

Again, extremes would lower the chances of life’s coming into being if a planet’s axis of rotation were inclined so steeply to the vertical (relative to its plane of revolution about its star) that the major portion of the planet would be in sunlight for half a year and in the dark for half a year.

And yet again, if a planet rotates very slowly, the days and nights are each long enough to allow undesirable temperature extremes.

If a planet is a little on the massive side, it may just happen to collect enough water to make its ocean a planetary one, with little or no dry land. Even if life then develops, it is not likely that technology will, and we are looking not for life alone, but technology as well.

In reverse, if a planet is a little on the nonmassive side and little water is collected, the world may be mostly desert, and life may at best form to only a limited extent and reach insufficient levels of complexity.

The atmosphere may not be quite right in some ways, and block off too much of the sunlight, or too little of the ultraviolet radiation. Or else the crust may not be quite right and there may be too much volcanic action or earthquakes. Or else the surroundings in near space may not be quite right and meteoric bombardments may be too intense for life to maintain itself.

None of these imperfections is very likely, perhaps. After all, among the planets of our Solar system, only two (Mercury and Pluto) have orbits that are significantly elliptical; only one (Uranus) has an enormous axial tilt; only two (Mercury and Venus) have very slow periods of rotation, and so on.

Yet although each one of the imperfections is unlikely in itself and may affect only one out of ten Earthlike planets, or fewer, all the various imperfections mount up.

Again, we might suppose (intuitively) that only one out of every two Earthlike planets is Earthlike in every important particular; that it has a day and night of reasonable length, seasons that do not go to unreasonable extremes, oceans that are neither too extensive nor too restricted, a crust that is neither too unsettled nor too geologically inert, and so on.

We might say that such planets are “completely Earthlike” or, better, simply “habitable.” In fact, we no longer have to specify that we are speaking of Sunlike stars, or of second-generation Population I stars, or of ecospheres. The term habitable would imply all that out of necessity.

If, then, one out of every two Earthlike planets are habitable, we have our eighth figure:

This sounds like a large number and, of course, it is, but it represents a measure of our conservatism also. This number means that in our Galaxy, only one star out of 460 can boast a habitable planet. What’s more, it is a more conservative figure than some astronomers would suggest. Carl Sagan, who is one of the leading investigators of the possibility of extraterrestrial intelligence, suggests there may be as many as one billion habitable planets in the Galaxy.