g2 KANSAS UNIVERSITY QUARTERLY. 
spite of the paucity of data the results obtained with this mill show 
conclusively that the most important element in the efficiency 1s the 
pump load factor.” 
The brake load of 2 lbs. corresponds to a useful pump load of 
740 ft. lbs. per stroke. This is somewhat less than the load of 
mill No. 11 and greater than the average pump load of this size 
of mill. It is seen that this 2 lb. curve is nearly a straight line for 
the usual jumping velocities of ro to 20 miles per Hounenwbuat is, 
the power of a power mill as wellas that of a pumping mill increas- 
es nearly as fhe first power of the wind velocity for a constant load 
of 2 lbs. 
The dotted curve D. K. is drawn tangent to these load curves and 
is the envelop of them. It is very nearly a parabola whose vertex 
is at the origin, with axis vertical. It is easily seen that this curve 
gives the relation between wind velocity and horse power for a con- 
stantly increasing load on the mill. The power of the mill would 
be greatly increased by some device for increasing the load of the 
mill it would then vary as the second power of the velocity instead 
of the first power. 
The curve A. B. gives the relation between wind velocity and 
horse power for the total energy of the wind which strikes the fans 
of this mill. That is, if this windmill could utilize all the energy of 
the wind which strikes the surface of its fans its power would then 
be given by the curve A. B. which is of the third degree. The ef- 
ficiency of the mill being the ratio of the horse power developed by 
the mill to the total horse power of the wind which strikes it, it is 
seen to be the ratio of the ordinate of any load curve to the corre- 
sponding ordinate of curve A. B. It is seen to decrease for a con- 
stant load as the velocity increases, and to increase for a constant 
velocity as the load increases. The maximum efficiency is the ra- 
tio of the ordinate of curve D. K. to the corresponding ordinates of 
curve A, B. 
It is seen that the useful work that a windmill will do when work 
ing under a constant load at all velocities is small, the horse power 
varying as the first power of the velocity. By some device for au- 
tomatically increasing the load as the wind velocity increases the 
power may be much increased at high velocities, and will then va- 
ry as the second power of the velocity. Even with this device for 
increasing power there is still a great difference between the power 
possessed by the wind and the power developed by the mill and 
this difference increases as the velocity increases. The question 
may be asked—‘‘Why does not the windmill utilize a larger amount 
