lands of the Missouri River. This, taken in connection with the fact
that the grit in the water cuts the pump plunger packing so fast that in
a short time the pump will not work up to its capacity, accounts for the
apparent small amount of power developed by this mill.

There has been some discussion of late in regard to the horse power of
wind mills, one party claiming that they were capable of doing large
amounts of grinding and showing a development of power that was
surprising to the average person unacquainted with wind mills, while the
other party has maintained that they were not capable of developing any
great amount of power, and has cited their performance in pumping water
to sustain his argument. My experience has has led me to the conclusion
that pumping water with a wind mill is not a fair test of the power that
it is capable of developing, for the following reasons:

A pumping wind mill is ordinarily attached to a pump of suitable size to
allow the mill to run at a mean speed in an 8 to 10 mile wind. Now, if
the wind increases to a velocity of 16 to 20 miles per hour, the mill
will run up to its maximum speed and the governor will begin to act,
shortening sail before the wind attains this velocity. Therefore, by a
very liberal estimate, the pump will not throw more than double the
quantity that it did in the 8 to 10 mile wind, while the power of the
mill has quadrupled, and is capable of running at least two pumps as
large as the one to which it is attached. As the velocity of the wind
increases, this same proportion of difference in power developed to work
done holds good.

St. Louis is not considered a very windy place, therefore the following
table may be a surprise to some. This table was compiled from the
complete record of the year 1881, as recorded by the anemometer of the