Cowboys, ranchers, and scientists have found thousands of solid nickel-iron meteorite fragments around the crater. The largest of these weighs 1,406 pounds. The smallest spherules and grains are almost or quite microscopic in size. (These tiny granules have been well known to scientists since 1905 in spite of current fables claiming that they are a recent discovery.) In the rim and on the plain outside the crater, large and small _shale balls_, composed of weathered meteoritic material, were found in considerable numbers in the early days. Along with many solid iron meteorites, shale balls have also been found at various depths in recent times by field parties from the Institute employing specially designed meteorite detectors. In the first two decades of the twentieth century, investigators sank (at great expense!) a number of shafts and drill holes in the interior and on the south rim of the crater, in unsuccessful attempts to locate the supposed “main mass” of the Canyon Diablo meteorite. Most authorities now believe, however, that the extremely high temperatures, developed at the time the Canyon Diablo meteorite penetrated into the earth, changed almost all of the gigantic cosmic missile into vapor. [Illustration: View of the interior of the Canyon Diablo crater showing the steep inner slopes of the huge basin.] No better example of an ancient meteorite crater has been found than this one near Canyon Diablo. The other craters listed in the table (even the two recently formed ones), while bearing resemblances to it, also show individual differences from it. Some, like Henbury, Campo del Cielo, and Haviland, are not single craters but rather consist of fields of craters. In these cases, the earth was struck not by a single large meteoritic body that held together right down to impact, but either by a “swarm” of meteorites traveling together through space or by the fragments of a large meteorite that separated into pieces shortly before it struck the surface of the ground. Again, the type of ground into which the meteorite strikes affects the character of the craters formed. As an illustration, the Wabar, Arabia, craters were not smashed out of sedimentary, horizontally bedded rock layers (as was the Canyon Diablo crater) but were formed in clean desert sand dunes. In this case, the crater rims are composed primarily of almost pure silica-glass formed by the fusion of the sand at the time of impact. It is not hard to imagine the terrific boiling and frothing up of melted sand and meteoritic material that must have accompanied the formation of the Wabar craters. Except for Podkamennaya Tunguska and Ussuri, the craters listed in the table were formed, as we have mentioned, a great many thousands of years in the past. Just how many thousands is a difficult question to answer, for all of our estimates must necessarily be made on the basis of _indirect_ evidence rather than on _direct_ observation.