x x, which corresponds to the zero of pressure on the dynamometer. At the instant of the beginning of the translation the apparatus shows an energetic pressure (a) on the dynamometer; this is the initial variable resistance caused by the inertia of the air against which the screen pushed. Very soon afterward the curve falls, announcing that the resistance of the air has diminished, although the motion of the disk had remained uniform. This is owing to the fact that the air had then partly acquired the motion of the screen. The pressure falls thus to the level b, which marks the constant resistance of the air during the whole remaining period of the translation.
Finally, when the apparatus is suddenly stopped, we see that the trace of the recording-lever is suddenly depressed at the point c; this is due to the variable terminal condition: it consists in the carrying of the screen forward by the column of air already set in motion by it. This negative resistance gradually ceases, and the tracer returns to zero.
We were not able to determine with this rough apparatus the absolute value of the resistance of the air corresponding to different instants of uniform translation, but we can readily see that there exist two variable states, of which one precedes and the other follows the constant resistance of the air. The studies of physicists have heretofore been directed only to the determination of the constant resistance corresponding to different velocities.
Second Series of Experiments.—Increase of the resistance of the air to the downward movement of the wing of a bird, caused by the horizontal translation of the bird.
In making the above determination, I have roughly imitated the construction of the bird, by reducing each of the wings to a thin and rigid plane of a 12 metre in length, and 110 a metre in breadth. These two wings are simultaneously depressed by the action of a spring.
A constant amount of work is thus employed for each blow of the wings. The translation of the artificial bird takes place in gliding along an horizontally-stretched iron wire. Two large wheels, one of them furnished with a crank-handle, move an endless cord parallel to the iron wire. The apparatus with wings is attached to this cord, and can thus be moved horizontally with greater or less velocities.
It is now necessary to determine with precision the velocity of translation and the duration of the depression of the wings. The graphic method gives readily these two measurements:
1. Measurement of the Velocities of Translation of the Apparatus.—This velocity is evidently the velocity of a point on the endless cord, to which the winged apparatus is attached. This cord passes around a little pulley, whose revolutions are counted and registered on a revolving cylinder by means of a lever which is worked pretty much as the lever in the registering apparatus of Morse's telegraph.
The little pulley which serves to measure the velocities is exactly
