Flaky particles of this size are easily carried along by a moderate wind. In some places it appears as if the dust were resting on an old land surface where no water could have been standing when it fell. There is really no room for doubt that it was carried several hundred miles by the wind. It must have darkened the sky at the time, and it must have settled slowly and quietly over the wide plains, covering extensive tracts with a white, snowlike mantle several feet in thickness. 
Fig. 4.—Tracks in the Volcanic Dust, probably made by a Crawfish. Reduced to 2⁄3 diameter. What a desolate landscape after such a shower! What a calamity for the brute inhabitants of the land!
Right here in McPherson County there was either a river or a lake at the time of the catastrophe. This is plainly indicated in several ways. In one place the dust rests on sand and clay, with imbedded shells of fresh-water clams. It is assorted in coarse and fine layers like a water sediment. Lowermost is a seam of very coarse grains. These must have settled promptly through the water, while the finer material was delayed. In another place it lies on higher ground, and here marks of sedges and other vegetation are seen extending up about a foot into the base of the deposit, from an underlying mucky clay. Bog manganese impregnates a thin layer just above the clay, indicating a marshy condition. Here also the material is somewhat sorted, but in a different way. It is ripple-bedded. The water was evidently shallow, if there was any water at all. A burrow like that of a crawfish extended down into the old clay bottom. On a slab of the volcanic ash itself some tracks appeared (Fig. 4). These were probably made by an individual of the same race in an effort to escape from the awful fate of being buried alive like the inhabitants of Herculaneum and Pompeii.
The shower must have lasted for a time of two or three days. I infer this from the nature of the wind changes, which are indicated by the ripples in the dust. These still lie in perfect preservation (Fig. 5), and may be studied by removing, inch by inch, the successive
