Gilbert saw two possible types of explanation for the crater: it could have been formed by something coming in – an impact – or by something coming out – a volcanic explosion. The best argument for a falling star was the meteoritic iron littering the surrounding desert. Gilbert calculated the odds of a crater forming in such a dense meteor field purely by chance as 800 to 1. If the crater had been clearly volcanic, then this might not have mattered. But though there were volcanoes nearby, the crater’s walls and floors were sedimentary rock, the same strata of sandstone and limestone from which the rest of the Colorado Plateau is built.
In a typically methodical manner Gilbert set out to test the alternatives through their implications. If there were a ‘star’ buried beneath the crater somewhere, then like Archimedes in his bath it would have displaced material that was there before. If so, there would be more material in the crater’s raised rim and its surrounding blanket of ejecta than was needed to refill the crater itself. But when, through painstaking surveying, Gilbert and his assistants compared the volume of the crater’s cavity with the volume of the rock that had been excavated in the catastrophe, they found that if the rim and ejecta were put back into the crater they would almost exactly fill it up; thus there was no evidence for the bulk of an added meteor below the crater floor. What was more, if a large iron meteorite did lie buried there it should have had a quite discernible magnetic field. But no such field was found. So Gilbert decided the crater had been formed by an explosion of steam, set off when deep volcanic activity had penetrated a subterranean aquifer; he placed Coon Butte in the family of anomalous volcanic craters called ‘maars’ (no relation). This hypothesis sat well with the natural inclination of the area’s uneducated shepherds: that the crater looked as though it had been formed by something exploding out of the earth, not by something falling into it.
Disappointed as he may have been – a maar is an interesting thing, but hardly a star – Gilbert still put his observations to good use. In his 1895 address as president of the Geological Society of Washington, published as ‘On the Origin of Hypotheses’, he presented the story of Coon Butte as a sort of moral fable on the correct way of approaching geology. To explain a novel feature, the geologist should first reason by analogy: what sort of thing is it like? The analogy might seem a distant one – a gaping crater in a desert is not very like the ‘raindrop falling on soft ooze’ to which Gilbert compared Coon Butte – but that need not matter. What matters is that there be a number of analogies, that they have different physical implications, and that those implications then be tested. This was Gilbert’s highly influential encapsulation of what was becoming the pragmatic cornerstone of geological science in America: a method of ‘multiple working hypotheses’ in which contradictory explanations were to be entertained simultaneously.
One of the disappointments for Gilbert in finding Meteor Crater to have been produced from within and not without was that he had hoped to use it as an analogy with which to bolster his theories about the moon. Everyone who wrote on the moon explained it by analogy to the earth; the problem lay in choosing the right analogy. In 1874 James Carpenter, a Greenwich astronomer, and James Nasmyth, an engineer whose father had been a landscape artist and whose own pictures of the moon had caught the eye of Prince Albert, published a wonderful illustrated book called The Moon: Considered as a Planet, a World and a Satellite.* Inside, spectacular photographs and prints of the moon are compared with similarly lit photographs of a range of other objects – an old man’s wrinkled hand, a desiccated apple, a cracked sphere of glass. The idea is to teach the reader’s eye new ways of seeing the moon and give his mind new analogies by which to understand it. (Their influence was long-lasting. Lowell used the desiccated apple in his books on Mars to demonstrate what happens when a planet dries up; the first post-Mariner textbook on Martian geology has very Nasmyth-like cracked glass spheres in it to demonstrate stress patterns.)
To make their case for the volcanic origin of the moon’s craters, Nasmyth and Carpenter created a scale model of what Vesuvius and the bay of Naples must look like from above and compared it with similar models of the lunar surface. Other lunar analogies on offer suggested that the dark expanses of the moon called ‘seas’ were in fact made of ice, or that they were the dried beds of seas now vanished. Charles Babbage, the pioneer of mechanical computing, elaborated on this idea with the notion that craters in these dried seas were in fact coral atolls like those studied by Darwin.
Gilbert rejected all these analogies, seeing the craters and larger basins and ‘seas’ as the marks left by planetesimals. His idea was that once the earth had been ringed by planetesimals – much as Saturn is ringed today – and that these had then coalesced into the moon; the last ones in had left the surface scarred. Lacking a natural earthly analogue for such cratering, Gilbert experimented with crater making himself, firing various projectiles into clay: he called the hobby ‘his knitting’ and found the results satisfactorily lunar. To those who objected to a geologist trespassing in the realms of astronomy, he defended his speculations in terms that could serve as the credo for astrogeology to this day: ‘The problem is largely a problem of the interpretation of form, and is therefore not inappropriate to one who has given much thought to the origin of terrestrial topography.’
Gene Shoemaker’s thinking on terrestrial topography, which would find application in the interpretation of form on the moon and beyond, took place in large part on the Colorado Plateau which Gilbert had known so well (indeed, he had given it its name). In 1948, twenty years old and with a Caltech degree in geology already behind him, Shoemaker joined the US Geological Survey and found himself working in southern Colorado. He discovered that he loved the landscapes of the south-west. He loved the pines, he loved the open spaces, he loved the great, vaulting skies. He stared up at the desert moon with wonder.
In the field, he did not have much contact with the rest of the world. But he did get the Caltech alumni newspaper, which revealed that experiments with captured V2 rockets elsewhere in New Mexico were reaching the very edge of the atmosphere. It was a revelation. ‘Why, we’re going to explore space,’ he later remembered thinking, ‘and I want to be part of it! The moon is made of rock, so geologists are the logical ones to go there – me, for example.’
Shoemaker kept his wild dream to himself – a decade before Sputnik there was little call for space age geology. The atomic age, though, needed geologists badly. Cold War strategy required that America develop reliable domestic sources of uranium, and the Colorado Plateau was thought likely to hold the reserves required. So in his first years with the USGS Shoemaker joined in the last great American mining boom; at the same time he started work on a Ph.D at Princeton and got married. He criss-crossed the Colorado Plateau from site to site, ‘half man, half jeep’, according to his wife, Carolyn, who often accompanied him. It wasn’t normal for geologists’ wives to come along on field trips, but the Shoemakers didn’t care. When they had children, the children came too.
All the while, Shoemaker kept thinking about the moon. He read everything there was to read on the subject, including Grove Karl Gilbert; he tailored his fieldwork to suit his extraterrestrial interests. It was this which led him to map diatremes in the Painted Desert’s Hopi Buttes. Diatremes are volcanic features, chimneys of magma that rise to the surface causing explosions, which throw out a lot of normally well-buried rock and comparatively little lava; they can create the low-lying craters called maars to whose number Gilbert had added Meteor Crater. As a uranium prospector, Shoemaker was interested in diatremes because the rocks they threw out when they cleared their throats might be from uranium-bearing strata. As a would-be lunar geologist, he was interested in them because their associated