We can understand the basic differences in content and structure between the terrestrial and the jovian planets using the condensation theory of the solar system's origin. Indeed, it is in this context that the adjective condensation derives its true meaning. To see why the planets' composition depends on location in the solar system, it is necessary to consider the temperature structure of the solar nebula.
As the primitive solar system contracted under the influence of gravity it heated up as it flattened into a disk. The density and temperature were greatest near the central protosun and much lower in the outlying regions. Detailed calculations indicate that the gas temperature near the core of the contracting system was several thousand kelvins. At a distance of 10 A.U., out where Saturn now resides, the temperature was only about 100 K.
In the warmer regions of the cloud, dust grains broke apart into molecules, and they in turn split into excited atoms. Because the extent to which the dust was destroyed depended on the temperature, it also depended on location in the solar nebula. Most of the original dust in the inner solar system disappeared at this stage, but the grains in the outermost parts probably remained largely intact.
The destruction of the dust in the hot inner portion of the solar nebula introduced an important new ingredient into the theoretical mix, one that we omitted from our earlier account of the accretion process. With the passage of time, the gas radiated away its heat and the temperature decreased at all locations, except in the very core, where the Sun was forming. Everywhere beyond the protosun, new dust grains began to condense (or crystallize) from their hotter gas phase to their cooler solid phase, much as raindrops, snowflakes, and hailstones condense from moist, cooling air here on Earth. It may seem strange that although there was plenty of interstellar dust early on, it was mostly destroyed, only to form again later. However, a critical change had occurred. Initially, the nebular gas was uniformly peppered with dust grains. When the dust re-formed later, the distribution of grains was very different.
Figure 15.6 plots the temperature in various parts of the primitive solar system just prior to the onset of the accretion stage. At any given location, the only materials to condense out were those able to survive the temperature there. As marked on the figure, in the innermost regions, around Mercury's present orbit, only metallic grains could form. It was simply too hot for anything else to exist. A little farther out, at about 1 A.U., it was possible for rocky, silicate grains to form, too. Beyond about 3 or 4 A.U., water ice could exist, and so on, with the condensation of more and more material possible at greater and greater distances from the Sun. The composition of the material that could condense out at any given radius would ultimately determine the types of planets that formed there.
Figure 15.6 Theoretically computed variation of temperature across the primitive solar nebula. In the hot central regions only metals could condense out of the gaseous state to form grains. At greater distances from the central protosun the temperature was lower, so rocky and icy grains could also form. The labels indicate the minimum radii at which grains of various types could condense out of the nebula.
In the middle and outer regions of the primitive planetary system, beyond about 5 A.U. from the center, the temperature was low enough for the condensation of several abundant gases into solid form. After hydrogen and helium, the most common materials in the solar nebula (as they are today in the universe as a whole) were the elements carbon, nitrogen, and oxygen. The most common chemical compounds were those containing those elements—specifically, water vapor, ammonia, and methane. As we have seen, these compounds are still the primary constituents of jovian atmospheres.
At temperatures of a few hundred kelvins or less, these gases condensed out of the nebula. Consequently, the ancestral fragments destined to become the cores of the jovian planets were formed under cold conditions out of low-density, icy material. The planetesimals that formed at these distances were predominantly composed of ice. Because more material could condense out of the solar nebula at these radii than in the inner regions near the protosun, accretion began sooner, with more resources to draw on. The outer planets grew rapidly to the point where they could accrete nebular gas, not just grains, and eventually formed the hydrogen-rich jovian worlds we see today.
With the formation of the four giant jovian planets, the remaining planetesimals were subject to those planets' strong gravitational fields. Over a period of hundreds of millions of years and after repeated "gravity assists" from the giant planets, especially Uranus and Neptune, many of the interplanetary fragments in the outer solar system were flung into orbits taking them far from the Sun. Astronomers believe that those fragments now make up the Oort cloud, whose members occasionally visit the inner solar system as comets. 
(Sec. 14.2) During this period many icy planetesimals were also deflected into the inner solar system, where they played an important role in the evolution of the inner planets.
A key prediction of this model is that some of the original planetesimals should have remained behind, forming the broad band known as the Kuiper belt, lying beyond the orbit of Neptune. (Sec. 14.2) In 1993 several such asteroid-sized objects were discovered, lying between 30 and 35 A.U. from the Sun and lending strong support to the condensation theory. Over 60 Kuiper belt objects, having diameters ranging from 100 to 400 km, are now known. (See also Interlude 15-2.)
In the inner regions of the primitive solar system, condensation from gas to solid began when the average temperature was about 1000 K. The environment there was too hot for ices to survive. Many of the abundant heavier elements, such as silicon, iron, magnesium, and aluminum, combined with oxygen to produce a variety of rocky materials. Planetesimals in the inner solar system were therefore rocky in nature, as were the protoplanets and planets they ultimately formed.
These heavier materials condensed into grains in the outer solar system, too, of course. However, they would have been vastly outnumbered by the far more abundant light elements there. The outer solar system is not deficient in heavy elements. The inner solar system is underrepresented in light material. Here we have another reason why the jovian planets grew so much bigger than the terrestrial worlds. The inner regions of the nebula had to wait for the temperature to drop so that a few rocky grains could appear and begin the accretion process, but the outer regions might not have had to wait at all. Accretion in the outer solar system began almost with the formation of the disk itself.
Very abundant light elements such as hydrogen and helium, as well as any other gases that failed to condense into solids, would have escaped from the terrestrial protoplanets or, more likely, were simply never accreted from the solar nebula. The inner planets' surface temperature was too high, and their gravity too low, to capture and retain those gases. Where then did Earth's volatile gases, particularly water, and the recently discovered ice on Mercury and the Moon, come from? The answer seems to be that icy fragments—comets—from the outer solar system, deflected into eccentric orbits by the jovian planets' gravity, participated in the meteoritic bombardment of the newly born inner planets, supplying them with water after their formation.
The myriad rocks of the asteroid belt between Mars and Jupiter failed to accumulate into a planet. Probably, nearby Jupiter's huge gravitational field caused them to collide too destructively to coalesce. Strong Jupiter tides on the planetesimals in the belt would also have hindered the development of a protoplanet. The result is a band of planetesimals, still colliding and occasionally fragmenting, but never coalescing into a larger body—surviving witnesses to the birth of the planets.
Most of the planetesimals left over after the major planets formed eventually collided with a planet or were ejected into the Oort cloud. Little solid material remained. But what of the gas that made up most of the original cloud? Why don't we see it today throughout the planetary system? In the outer solar system, some (but not all) of that gas was swept up into planets. But that did not occur in the inner regions, where the terrestrial protoplanets never became massive enough to accrete such light material. Instead, the newly formed Sun took a hand.
All young stars apparently experience a highly active evolutionary stage known as the T Tauri phase (Figure 15.7; see also Chapter 19), during which their radiation and stellar winds are very intense. Any gas remaining between the planets was blown away into interstellar space by the solar wind and the Sun's radiation pressure when the Sun entered this phase, just before nuclear burning started at its center. Afterward, all that remained were protoplanets and planetesimal fragments, ready to continue their long evolution into the solar system we know today.
Figure 15.7 (a) Strong stellar winds from newly born stars are responsible for sweeping away any dust and gas left over from the star formation process, (b) leaving only planets and planetesimals behind.
