*POPULAR SCIENCE MONTHLY.*

for a general estimate, which is made here in the absence of a better one.

If we say that this western country experiences a dust storm twice a year, we do not rate this work too high. During two such days the velocity of the wind for the lowest mile in the atmosphere will average, at least, thirty miles an hour, and the total wind movement will be 1,440 miles. Of course, there are many sheltered places where such winds will not be felt. The territory including the west two thirds of the Dakotas, of Nebraska, Kansas, Oklahoma, and Texas, and extending west to the Pacific Ocean, contains about 1,000,000 square miles of open land, allowing one third the area for mountains. For the time of the storm the atmosphere over this area may be regarded as a current of air 1,000 miles long, 1,000 miles wide, and a mile high, containing 1,000,000 cubic miles of air. If the lowest 200 feet in this current carry a load of 20,000 tons to the cubic mile, and if the remaining 5,080 feet carry 100 tons per cubic mile, there will be 853 tons of dust to each square mile of the whole area of the current, making a total of 853,700,000 tons transported a distance of 1,440 miles, or, if the expression be permitted, 1,229,342,400,000 *mile-tons* of transportation will be performed.

Comparing this with the quantity of work performed by the water of the Mississippi drainage system, we find that the latter is three hundred and thirty times as great. The Mississippi carries annually 406,250,000,000 tons (estimate by Humphrey and Abbott) of sediments a distance of, say, 1,000 miles, performing 406,250,000,000,000 *mile-tons* of transportation. The ratio of the atmospheric transportation in the West and the aqueous transportation in the Mississippi basin is then 1:330.

If the above estimates have any significance at all, it is to the effect that in this country the work of the atmosphere is less than the work performed by meteoric waters. So far the inference is in full accord with the well-grounded general consensus among geologists.

Care has been used to not overrate any of the factors entering into the estimates. In sandy regions a considerable amount of transportation is effected by pushing by the wind on the loose surface material. This part of the work of the wind is here entirely neglected. It is known, too, that even on calm days the atmosphere carries an appreciable load of dust, and this has not been taken into account, though it operates for the remaining three hundred and sixty-three days of the year. It is, therefore, possible that the general estimate is too low. But it is not necessary to have recourse to such a supposition alone to show that, locally, work by the atmosphere may exceed the work performed by the water. The observed average wind velocities in these