Fundamentally, any debris that did not originate from Earth that survives a descent through Earth's atmosphere is called a meteorite.  Only in the 19th century did astronomers recognize that such objects - extraterrestrial rocks - could exist.

Why? Why would the concept of a rock falling from interplanetary space seem impossible before the 19th century? The answer to this question illustrates much about the process of science.

Prior to Galileo's work in the early 17th century, all  physicists and astronomers (so-called  "natural philosphers") and any other thinkers of profound thoughts (theologians, philosophers, mathematicians) believed in a universe in which the Earth was at the center, while the Sun, Moon, planets and stars all revolved around the Earth.  In that universe, everything was made of five elements: earth, air, fire, and water, all of which existed down here in the "terrestrial realm," and the aether, which made up all celestial objects.  Since a rock from the sky would have to be a piece of "earth" falling from the celestial realm, and since by definition objects must be made of aether if they are in the celestial realm, a rock from the heavens could not exist.  It violated the laws of physics, as understood.

After Galileo, we understood that the planets were likely made of earthlike stuff.  But the universe consisted only of stars, planets and a few moons. Only in the mid-18th century were comets understood to be objects orbiting the Sun; but their nature was not understood, and nobody considered the likelihood of one hitting the Earth.  In fact, most astronomers believed the solar system must be extremely stable, almost changeless since it was created, presumably by a friendly deity, about 6000 years ago.

But all that changed in the 19th century, with the discovery of a multitude of asteroids and the growing understanding that a ring of asteroids, the asteroid belt, encircled the Sun.  In addition, many astronomers believed that the asteroids were fragments remaining from a collision that destroyed a planet that once existed at 2.8 AU, as predicted by the Titius-Bode formula; thus, the idea that  objects  orbiting the Sun could be destroyed, creating countless fragments in new orbits finally creates all the conditions necessary -- a reservoir or rocky material orbiting the Sun, a mechanism for making small pieces of debris and sending them into new orbits -- permits the possibility that a rock could fall from the sky.

This illustrates a fairly dramatic -- though very slow (300+ years) -- paradigm shift in this area of astronomy.

Most such objects are fragments of asteroids. And only in the 1980s did astronomers identify meteorites that are unambiguously from Mars (16 meteorite pieces) and the Moon (also, 16 meteorite pieces).

Prior to the 1980s, most meteorites were finds - unusual looking rocks found usually accidentally - or falls - rocks that were observed to fall from the sky and collected immediately.  Beginning in the 1980s, expeditions to Antarctica have produced more meteorites than the total number collected prior to these meteorite hunts.

Good information and pictures of meteorites can be found at Meteors, Meteorites and Impacts, this link, and The Meteorite Gallery.

Classifications of meteorites by inspection

Irons - meteorites composed almost entirely of nickel and iron.  Pure metal.  Most finds are irons; this is likely because pure iron rocks are odd objects to find in many locations. Picture.  Only 4% of the known meteorites are Irons.

Stones - meteorites composed almost entirely of rock, or silicates (silicon + oxygen, combined with other elements).  Most falls are stoney meteorites; talso, most Antarctic meteorites are stones. This large abundance of stones suggests that most of the source material for meteorites are stoney objects.   Picture.  Fully 95% of the known meteorites are stones. Of these, 5% are "chondritic" (see below).

Stoney-Irons - mixtures of iron-rich and silicate-rich material in one object. Only 1% of the known meteorites are stoney-irons.

Meteorite Ages

Virtually all meteorites have ages between 4.48 and 4.56 BY.  The average age of meteoritic material is 4.54 BY.  The very oldest meteoritic material is dated at 4.5578 +/- 0.0009 BY (or 4557.8 +/- 9 MY).   The only objects with younger ages are igneous (volcanic) rocks that are identified as having originated on the Moon or Mars.

Classification of meteorites by Process of formation

The fact that we have irons, stones, and stoney-irons indicates that there must be a process that was active somewhere at sometime that was able to generate these three types of meteorites.  Almost certainly, "when" is answerable as "sometime during the history of the solar system, i.e. with the last 4.56 BY" while "where" is answerable as "somewhere within our solar system."  Can we do better than this?

First, we need a process and for that, we need to consider the abundances of the elements and think about what objects in the solar system and elsewhere are made of.

Sun - 98% H + He, 2% all other elements
Mercury - iron-rich
Venus - 25% iron/nickel, 75% rock
Earth - 25% iron/nickel, 75% rock
Mars - 25% iron/nickle, 75% rock
Asteroids - most appear rocky, likely contain iron/nickel
Jupiter - 3% iron/nickel/rock, 97% H + He
Saturn - 3% iron/nickel/rock, 97% H + He
Uranus - 60% iron/nickel/rock, 40% H-rich compounds (methane, ammonia, water), some He
Neptune - 50% iron/nickel/rock, 50% H-rich compounds (methane, ammonia, water), some He
Pluto, KBOs - mostly water ice

other stars - 97-98% H + He, 2-3% all other elements
galaxies - 97-99% H + He, 1-3% all other elements
universe - 98% H + He, 2% all other elements

Thus, it is clear that metallic and rocky objects are the dominant material in the inner solar system and that meteorites most likely come from the inner solar system. We them might conclude that they are fragments of existing objects (rocks kicked off the Moon, Mars, Venus, Mercury, asteroids) or are pieces of material that never formed into larger objects.

Assuming the building blocks of the terrestrial planets were chunks of metal-rich and silicate-rich rocks, as we build larger objects (called planetesimals), what happens to the insides of these objects?  The self-gravity of the objectd increases as the size increases. Thus, material at the center is squeezed.  What happens to material if you increase the pressure on it, but don't necessarily change the volume? The temperature must increase.  Even without radioactive heating, the inside of the Earth would be warmer than the outside because of this squeezing effect of gravity.  If the inside gets warm enough, the solid rock and iron becomes soft; first it becomes "plastic" (i.e., deformable), then it liquifies.  And what happens in an environment dominated by gravity if the material is a liquid made of many different types of materials?  The heavy (or dense) materials settle out to the bottom, the light materials rise to the top. This process of separation of dense from less dense materials is called differentiation.

If an object made of evenly mixed metallic and rocky materials is able to differentiate, what will happen?  The metals will settle to the "bottom" or core of the object while the rock will rise up to the crust.  In between, we'll have different degrees of mixing.  This is what happened to Earth: we have a iron-nickel  core surrounded by a rocky mantle and crust.  Crustal rocks are much less dense than the mantle rocks, while all the rocky material is less dense than the iron core.  Thus, Earth is a differentiated object.

So what is the best way to create irons, stoney-irons, and stoney meteorites?

Such objects must be fragments of differentiated parent bodies.  Such parent bodies were large enough to become molten and permit differentiation within, before a collision destroyed the parent bodies.  Fragments of the core became candidates to become iron meteorites; fragments of the mantle or crust became candidate stoney meteorites.

The best candidates for meteorite parent bodies are asteroids.  Asteroids all have fairly distinct "signatures" in the form of the spectrum of reflected sunlight from their surfaces.  Some objects are bluer, others, greener, others redder.  The relative colors provides the asteroidal "reflectance spectrum" signature.  We can do the same thing with meteorites: measure the colors of reflected sunlight.

What do we find?  We can match many meteoritic spectra with those of asteroids and thus can determine the part of the asteroid belt wherefrom  the meteorite likely was ejected.

Based on matching reflectance spectra, this object is believed to be a fragment of the asteroid Vesta.

What kind of object is "large enough" to become molten and permit differentiation, simply from gravitational pressure? Not the moon. Perhaps not even Mars.  Venus and Earth, probably are.  So, does this indicate there must have been an Earth-sized planet that was destroyed? No.  But it does indicate there must be additional mechanisms to heat up smaller objects.  The most likely mechanism is radioactivity.

If an object has radioactive material inside, the heat from radioactive decay will heat up the object.  Since the outside of the object is cool, radiating heat off into space, the internal heat will be conducted to the surface and the object will attempt to cool off.  In a small object (smaller than a km in diameter), the rate of heat production from radioactivity is slower than the rate at which heat is conducted to the surface.  So, these "small" objects will never heat up, become molten, and differentiate.  But larger objects (a few km in diameter) have a harder time cooling off (they have a smaller surface to volume ratio: 4 pi R2/ (4/3 pi R3) = 3/R), so they heat up more. These "large" objects could differentiate, given a good heat source.

The long-lived radionuclides that power plate tectonics in the Earth - uranium, thorium and potassium - produce heat too slowly to heat up an asteroid-sized object. But the radionuclide 26Al , which decays to 26Mg with a half-life of 700,000 years, could have heated up asteroid-sized objects during the first few million years of solar system history.  Once the asteroid differentiated, a collision any time during the last 4.5 BY would produce fragments that could become iron or stoney meteorites.

Carbonaceous chondrites

Most stoney meteorites are undifferentiated!  How can we tell?  Some stoney meteorites are pure stone; they are from a differentiated  parent body.  But most stoney meteorites contain silicate grains intermixed with metal-containing grains.  These objects have been much less processed than stoney meteorites than contain no metals. Picture.

Some undifferentiated meteorites contain  small spherical blobs known as chondrules.  "Chondrules" comes from the Greek word for a grain of seed, "chondros."   All chondrites (meteorites containing chondrules) have radioisotopic ages of 4.56 BY.

Some chondrites contain carbon-bearing chondrules. These objects are known as carbonaceous chondrites (CCs).  The material in the chondules is all very fragile.  Such carbon compounds are easily destroyed by heat and pressure.  IIf these chondrules had ever been exposed to much heat or pressure, they would have changed: some compounds within the chondrules would have been destroyed - in the same way that an egg changes when you heat it, even at very moderate temperatures - and others would have  lost their isotopic identities.  These carbonaceous chondrites are the most important meteorites because they are the oldest, least changed pieces of material from the first moments of the birth of our solar system. Everything else in the entire solar system (except perhaps pieces of comets) has lost some or all of the interesting evidence that would tell us about the origins of the solar system.  The carbonaceous chondrites are the smoking guns, containing evidence from the solar system's original moments.  Like regular chondrites, all CCs have radioisotopic ages of 4.56 BY.

Chondritic Abundances

Under the assumption that chondrules contain the best, least changed information about what the original material out of which the solar system was made, we can measure the abundances of all materials in chondrules to determine what the Sun and planets were made out of.    A comparison of chondritic with solar abundances (see Figure 1 in chondrules) reveals a nearly perfect correspondence between the two.  You can see from the figure that oxygen (O) is most abundant, followed by carbon (C), iron (Fe), magnesium (Mg) and nitrogen (N). The important result is that each element has the same abundance (say, in parts per billion) in chondrites as in the Sun.  The only exceptions are Boron (B) and lithium (Li) which are destroyed in the nuclear fusion processes in the Sun (hence they are depleted in the Sun relative to chondrites) and C, N and O, all of which are produced in nuclear fusion reactions in stars and thus are enhanced in the Sun relative to chondrites.    The extremely close correlation suggests that, in contrast to rocks on the Earth or Moon, chondrites have been unaltered (by heat, pressure, gravitational effects, chemical effects) since the Sun and solar system formed.  Thus, we take chondritic abundances to represent the relative amounts of the elements out of which the Sun and planets were made.

The Allende meteorite

The Allende metorite, a carbonaceous chondrite,  fell in Mexico at 1:05 AM on February 8, 1969.  Witnesses observed a fireball in the sky. Two tons of meteorite pieces were recovered (pictures of chunks of Allende, for sale; a thin section slice of Allende showing chondrules).  Allende fell, and pieces were collected, just when NASA funded research labs were gearing up to study sample returns from the moon (the first moon landing was on July 19, 1969).  So, state-of-the-art labs got pieces of a pristine carbonaceous chondrite at the perfect time.

Meteoritic researches found 26Mg in some chondrules in the Allende meteorite.  The 26Mg is found where normally, chemically, one would find Aluminum, not Mg.  But normally, Aluminum is 27Al.  Thus, it is clear that the 26Mg came from the radioactive decay of 26Al.  (One also comes to this conclusion from recognizing that the ratio of 24Mg/26Mg is much smaller in this meteorite than in almost all other rocks in the solar system.)

26Al has a half-life of only 700,000 years.  If we made some 26Al, 27Al and 24Mg (normal magnesium) and let it sit around as dust for 5 million years (7 half lives) before making chondrules and mineral grains, most of the 26Al would be gone. If we then made a rock out of the mineral grains, all the 26Mg would end up in minerals where Mg belongs, not where Al belongs.   Thus, the 26Al must have been locked into the aluminum-rich mineral grains fairly quickly after it formed.  The best estimates (which also use discoveries of the daugher products of the short-lived radioisotopes 41Ca, 53Mn, 60Fe, 129I, and 207Pd) are that all chondrule formation in the solar system ended within 7 My, and probably within 2 MY, of the beginning of the formation of the solar system.

Once locked in,  the decay would have rapidly heated the rock in which it was embedded, provided the large, surrounding rock was a few  km in size. (A 1.2 km diameter rock could heat up to 1800 K simply from the decay of  26Al, assuming it was made of 1% Al, as is typical of most materials).  So, our big rock with 26Al heats up; the heated rock becomes soft, even molten, the heavy elements (iron, nickel, palladium, lead, etc.) are pulled by gravity harder and thus sink to the center.  The ligher elements are pulled less hard and thus are nudged out of the way by the heavy stuff.  So the heavy stuff sinks, the light stuff rises.  After a few million years, the 26Al is pretty much spent so not much more heat is produced.  The rock cools off, solidifies; the core is nearly pure iron, the outside pure rock.  At some later time, two of these rocks collide.  The debris from the collision has pieces of the core which will become iron meteorites if they every hit the Earth; pieces of the outside will become rocky meteorites.

Thus, the detection of clear evidence that 26Al existed in the early solar system supports our claim that this was the heat source that enabled small asteroids to differentiate.  But, once an object heats up enough to differentiate, the signature of 26Al (26Mg  found where 27Al should be) is erased, too.  Thus, the chondrule in Allende contains the evidence for the presence of 26Al, but we conclude that this particular chondrule never heated up.  It was never in a large parent body.

Many of the small inclusions in Allende are even thought to contain small mineral grains that were formed in the outer layers of dying stars.  Such grains are known as presolar grains.  See Shifting Sands of Presolar  History  for more information.

The Murchison Meteorite

The  Murchison meteorite is also a carbonaceous chondrite.  It was observed to fall, in Australia, on September 28, 1969.  About 100 kg of material has been found. Murchison has a radioisotopic age of 4.56 GY.  Like Allende, pieces of the Murchison fall were collected immediately and so were relatively uncontaminated by weathering processes on Earth.

Murchison contains 12% water, an incredibly high water content for a meteorite.  Thus, it is believed to be a fragment of a comet rather than an asteroid!  Incredibly, Murchison also contains 92 amino acids (so far identified), of which only 19 are found on Earth.  Five of these 19 are found in living things on Earth.  The other 14 are not and are quite rare on Earth.   In addition, Murchison contains adenine, guanine, uracil, cytosine, and thymine (all five cross-linking bases in DNA and RNA).

Thus, at least  73 of these amino acids must have formed in elsewhere.  Thus, it is likely that the other 19 also formed elsewhere.  But do we have better proof of that?

Amino acids are fairly simple molecules  They are all composed of carbon, hydrogen, oxygen and nitrogen (the four most abundant elements in the universe, excluding helium).  In life on Earth, the most complex amino acid contains 25 atoms (Lysine = C6H15O2N2), the simplest contains but 13 atoms (Alaine = C3H7O2N).

Amino acids also have an interesting property in that they are either "right-handed" [d-amino acids] or "left-handed" [l-amino acids].  (A few, like glysine have no "handedness.")   Amino acids produced nonbiologically should have no preference for either form (they are "racemic").   Yet, all amino acids involved in life on Earth are left-handed.  The reason for this anomalistic preference is unknown, but living things select and manufacture only left-handed amino acids.  l-amino acids will make proteins usable by living cells; however a mixture of l- and d-amino acids  will not. [Note that a "right-handed" sugar would taste the same but would not be metabolized; thus, this would be a great dietary product, but so far, none has been successfully manufactured.]

What about the amino acids in Murchison?  To a very large degree, the left- and right-handed amino acids in Murchison occur in equal abundances.  In addition, the amino acids in Murchison have more deuterium (heavy hydrogen; hydrogen with 1 proton and 1 neutron) than terrestrial abundances (all objects in the solar system appear to have unique signatures in their relative abundances of hydrogen to deuterium) and have different isoptic signatures in their ratios of 12C/13C.

If the Murchison amino acids were terrestrial contaminants (which would be hard, since the meteorites were collected almost immediately upon falling to Earth), they should be only left-handed.  They are not!  Therefore, the amino acids found in Murchison must have formed elsewhere, either within this meteorite (or its parent body) or prior to incorporation within the meteorite parent body.

More detailed, more recent work on Murchison reveal that the amino acids have slight asymmetries, with measurably more left- than right-handed molecules. This slight asymmetry has now been shown for both life-bearing amino acids (glutamine, asparagine, proline, leucine, alanine; with l/d proportions from 64:36 to 83:17).and for four of those foreign to Earth and life on Earth (an l-excess from 2 to 9%).  Thus, it appears that a slight left-handed preference exists in the Murchison amino acids and that nature has a mechanism for selecting, preferentially, left-handed amino acids.

What does this mean?