September 17, 1969, found me in one of many white buildings on the campus of the newly created NASA Manned Spacecraft Center (MSC, Johnson Space Center) in Houston, Texas. I had been summoned there by a telegram from the curator of the Lunar Receiving Laboratory, to pick up my 10-gram allocation of lunar surface material that had been collected by the Apollo 11 astronauts during their lunar touchdown on July 20th.

I was approved as a min-pet principal investigator (PI). I am a hard-rock petrologist (Virginia Tech, MIT). The petro-root in this word means rock; thus petroleum = rock oil. Petrography is the descriptive end of the business. Once I heard a business-school guy tell his girlfriend that soft-rock geologists work for oil companies and drive Cadillacs, while hard-rock geologists are lean and hungry. I liked that, it made me quietly proud of the choice I had made. In graduate school I had focused my hard-rock passion on stony meteorites when I learned they are the oldest rocks to be found, meaning they might contain information about the origin of the Earth. Beginning in 1958, I spent my career (with a few digressions) at the Smithsonian Astrophysical Observatory (SAO) in Cambridge, Massachusetts. My boss in the Apollo era was Fred Whipple, the first Director of the SAO, who encouraged me to set up a meteorite laboratory at the Observatory.

At first I worked alone, but knew I needed help once I learned that I had been accepted to the lunar sample program. I recruited two freshly-minted petrology PhDs, John Dickey from Princeton and Ben Powell from Columbia; and I enlisted Ursula Marvin, a mineralogist and X-ray diffraction expert, who was already at SAO. And I hired Janice Bower, a remarkably adept and adaptable librarian, to train up to operate and service the electron probe microanalyzer that we were to acquire (Figs. 6 and 7). Electron probes perform microanalyses of the chemistry of mineral samples by analyzing the x-rays created when they are bombarded by an accelerated electron beam in a vacuum. Janice trained for the purpose at Wentworth Institute of Technology in Boston.

I boarded an Eastern Airlines flight to Boston with a stopover in Washington DC. In the plane I found a half-dozen or so other scientist friends and acquaintances, who were also flying home to Boston or DC with lunar samples they had picked up at MSC. We were all psyched and had much high-spirited conversation in the plane's aisle. We were probably obnoxiously loud. All this made me very warm, so I tore off my jacket and stuffed it in an overhead bin. Not until much later did I realize I had parted company with those two precious vials.

Once home, I displayed the lunar samples I had to my family, then to our neighbors and their kids. The next day I held a display in our lab for personnel at the Harvard Observatory and their families.

Everyone was excited and curious about the lunar samples, never mind their nondescript gray appearance (Fig. 12).

When my group was finally alone with our lunar samples, we examined them under a binocular microscope. We found to our surprise that our soil sample was not just a pulverized mineral dust, as I had pessimistically assumed it would be. The coarse-fines sample (10085,24) consisted of tiny miniature rocks, each with a distinctive texture and assemblage of minerals (Fig. 13). We had hundreds of separate lunar samples in our tablespoonful of that sieve fraction! So we set to work to study as many of those rock-ettes as we could in the time we had. Which was not much: NASA had decreed that on January 5, 1970, all of the lunar sample investigators were to convene at the Albert Thomas Convention Center in Houston, Texas, and present the results of their studies. Just 109 days after I picked up our sample.

Petrographers study rocks in thin sections, slices of rock about 30 micrometers thick (a human hair is about that diameter) mounted on glass microscope slides. These slivers are created by a rather tedious process of diamond sawing, grinding, and polishing on the surfaces of spinning laps (Fig. 14). NASA would make thin sections and send them to us, but these would be few and long in coming. So we prepared our laboratory to make our own thin sections. Soon our lab was a beehive of activity, as we sectioned, photographed, and analyzed 1,676 coarse-fine soil particles in batches on slides in our electron microprobe.

As data accumulated, it became clear what the Mare Tranquillitatis regolith consists of (Fig. 15). About half of our fragments were soil breccias, which is to say volumes of fine lunar soil that had been crammed together and lithified by impacts. You could say that destructive impacts giveth lunar soil and they taketh it away by consolidating it into breccia (an Italian word meaning "broken").

About 5% of the particles were glasses: volumes of rock or soil that had been melted by the energy of impacts and which then cooled rapidly by radiation into space. The amount of impact-melted glass in the Apollo samples surprised everyone.

Another 40% of the soil was particles of crystalline igneous rocks. It might seem they would all be similar representatives of the solidified lava that lies beneath Tranquillity Base, so the job of describing them would be simple, but that was not to be the case. Studies have shown that the debris from a cratering impact can travel great distances, so the soil particles that have collected at any point have come from many far-flung sources. Most of the igneous particles consisted of hardened basaltic lava (it had long been evident that the lunar maria must have been filled with lava); but there were many varieties of lava, differing in chemistry and texture.

However, another 3-4% of the igneous particles were something quite different and unexpected. They were white and consisted principally of the mineral anorthite (CaAl₂Si₂O₈). This type of igneous rock (anorthosite) is rare on Earth; an important anorthosite occurrence is in the Adirondack Mountains of New York. The word was first spoken in our lab when John Dickey was reading the microprobe analysis of a colorless glass droplet in our collection: "That's an anorthosite composition," he said (Fig. 16). No previous work had predicted that the Moon would contain a rock type so rich in aluminum and calcium as anorthosite.

This puzzled us. It is not enough for scientists to describe things, the whole point is to understand them. First, where had the light-colored anorthositic particles come from? That question was not so hard to answer. The Apollo lander Eagle had set down on the edge of dark, mare basalt-filled Mare Tranquillitatis, only about 50 km from the beginning of the whiter (more reflective) lunar highlands or terrae. Impacts on the terra regolith surely would have scattered some of it over into the mare regolith. This must have been the source of the anorthositic fragments in our sample (Fig. 17).

Second, did this unexpected composition hold for all of the terra rock that covers most of the Moon, or just that corner of it? Terra make up more than 80% of the lunar surface, counting both the near- and farside hemispheres; if all that rock is anorthositic, then this unexpected igneous rock type must comprise a significant fraction of our planet’s natural satellite.

An earlier robotic mission to the Moon, Surveyor 7, appeared to hold the answer to that question. Surveyor 7 carried a device called the alpha-scattering surface analyzer, which measured the chemical composition of the surface on the ejecta blanket of the large crater Tycho, whose dramatic rays splay across the face of the Moon. Tycho is in lunar highlands material 1,700 km from Tranquillity Base. The analysis it performed was consistent with an anorthositic composition there too. However, uncertainties in backscattering analysis, in particular its inability to distinguish between calcium and potassium, leave the analysis somewhat unsatisfying.

Generalizing boldly, a highly-reflective anorthosite-rich crustal layer seems to cover the whole Moon, atop whatever lies beneath, except where giant impacts have blasted holes through it which filled with basaltic lava.

Can the thickness of this layer be estimated? The principle of isostasy allows such an estimate to be made. Rock has plastic properties over long periods of time, meaning a heavy load on the crust will push subcrustal rock aside, making room for the load to sink. Because of isostasy, a mountain range can stand only if the mountains are lighter (less dense) than the rock beneath them, so they can "float" in it. The light (specific gravity 2.9) lunar crust can float 3 km above the denser interior material (Moon's overall specific gravity = 3.35), as it does, only if the crustal thickness is about 25 km. A lot of anorthosite!

Third, where did this layer of rare anothite-rich rock come from? The crystallization sequence for igneous rocks (experimentally determined) predicts that the mineral olivine crystallizes from cooling molten rock first, then pyroxenes and calcic feldspar, then feldspar richer in sodium. Olivine is dense, and it would tend to sink to the floor of a magma layer during crystallization. Crystallizing calcic feldspar (anorthite) is lighter, and under some circumstances it would tend to float rather than sink, accumulating at the top of the body of magma. This seems to be the only possible explanation for an anorthositic crust on the Moon (Fig. 18).

How much magma would have been needed to form the crust by crystal fractionation? Assuming a plausible bulk chemical composition for the Moon, it turns out most or all of the Moon  must have melted in order for 25 km of anorthite crystals to float to the top! I coined the term "magma ocean" to describe this huge molten mass, and the term has caught on.

I persuaded my group that this anorthosite story was what I (as our principal investigator) should stress in my talk at the impending Lunar Science Conference.

The conference was a very colorful experience. It began with a cocktail party for participants at Rice University. NASA had told us that we should not release our findings to the public; save them instead for a special issue of Science magazine that was to be dedicated to the first lunar sample reports. Most of us took that to mean we should not even spill the beans to other research groups (though this was not NASA's intent), so the party was an amusing cat-and-mouse game in which most of us were trying to find out what our colleagues had learned without revealing our own discoveries.

The first day of the conference (Monday, January 5th, 1970; by then the Apollo 12 mission had also been executed) was dedicated to results from the elite research groups, such as Jerry Wasserburg's self-styled "Lunatic Asylum" at Cal Tech. These workers determined Sr87/Sr86 ratios in several of the basaltic mare samples the astronauts collected, corresponding to an age of 3.65 billion years, about a billion years after the time when the solar system is understood to have formed.

Tuesday night there was a conference banquet in a more staid setting, where the speaker was Cambridge University astrophysicist Fred Hoyle. Among other things, he said this: "Our judgment of what were the significant issues in past times differ tremendously from contemporary judgment. In the middle of the 18th century the English celebrated victory at the end of a seven- year European war. Someone had the idea of getting George Frederick Handel, who was then an old man, to write a suite of music to celebrate the famous victory. As astute commentator, two centuries later, remarked that the whole meaning and purpose of this seven-year war had now been lost, and that in retrospect it appeared like an elaborate device to get old man Handel out of retirement and to get him to write his "Music for the Royal Fireworks". This metaphor was an introduction for his main message, which was that some day we may realize Apollo's most important contribution was the majestic view the spacecraft gave us of the whole pale blue Earth in the firmament (an allusion to the “Earthrise” photograph taken by astronaut William Anders on the Apollo 8 Mission, 1968). Perhaps, he thought, that view might bring home Earth's fragility to its inhabitants, and inspire them to take better care of their home.

If only.

My presentation (6-8) for the SAO group came at 5:20 PM Wednesday afternoon, the last talk of the day, when everyone was tired and looking forward to dinner. But the anorthositic-crust story interested people, and no one else had told it. Several groups had noticed and described the anorthositic particles in the lunar soil sample, but none pursued their meaning. Other min-pet groups tended to focus on the larger rocks that had been collected, fragments of the titanium-rich basaltic lava that filled Mare Tranquillitatis. Our picture of crustal formation by anorthite flotation during crystallization of a magma ocean has become widely accepted. The Moon is now widely believed to have been created by a giant planetary collision, in which a Mars-sized body spalled off a disk of molten debris from the Earth which then accreted into an orbiting Moon (9). Such a process easily could have been energetic enough to account for the initial molten state of the Moon.

Everybody's reports were in the framework of an igneous history of the Moon. This answered a burning question the media had been asking us, was the early Moon hot or cold? So little had been known of the character and composition of the Moon prior to Apollo 11 that this was not really known. The betting had been on a cold Moon, because that was the opinion of Harold Urey, a highly-respected Nobel-prize-winning chemist. In order to distance himself from his earlier work on the atomic bomb, Urey had interested himself in the chemistry of the solar system. On the basis of much hard thinking but little observational evidence he formulated a model to account for the solar system. In this model many cold "primary objects" accreted, too small to be heated significantly by decay of radioactive nuclides they contained; and all but one of these subsequently collected to form "secondary objects" (the terrestrial planets). The exception was Earth's Moon, a surviving primary object which Earth captured into orbit. But the Apollo 11 evidence of an igneous Moon laid this concept to rest.