One interesting outcome of his numerous experiments was the Hargrave Kite, now more familiarly known as the box kite. A good example of his kites is the type shown in Fig. 35. This consists of two arched biplanes mounted tandem on a backbone, or connecting framework. The kite floats steadily, and was thought suitable for the body of a flying machine to be driven by an engine and propeller. Thus meteorology is indebted to aëronautics for its most useful kite. [Illustration: FIG. 35.—HARGRAVE’S KITE.] A very novel and interesting type of aëroplane model was tested by Mr. Horatio Phillips in 1893. After careful preliminary experiments with various forms of curved “sustainers,” or lifting surfaces, tested in a wind tunnel, to determine which were most suitable wing forms, he finally constructed the flying apparatus shown in Plate XIV. This consisted of a compound aëroplane composed of many superposed narrow curved slats, the whole resembling an open Venetian blind. These curved blades, or sustainers, measured 12 feet long, 1.5 inches wide, 2 inches apart, and were held in a frame sharpened to cleave the air with slight resistance. The entire aëroplane spread 136 square feet of lifting surface, and was mounted on a truck as shown, carrying a steam engine and boiler, to actuate a two blade propeller 6 feet in diameter. The whole apparatus weighed 330 pounds, to which a dead load was usually added, and ran around a circular wooden track 628 feet in circumference, being tethered at the center, as in Tatin’s experiment. The apparatus readily lifted itself, when running at a speed of 28 miles an hour, and carried at the rate of 72 pounds per horse power, the added load weighing at times nearly one fourth that of the machine itself. The ultimate purpose of the experiment was to prepare the way for a one-man aëroplane like that shown in the lower part of the figure. This latter model actually carried a man across a field in 1904, but was found defective in longitudinal balance, because perhaps of its inadequate horizontal rudder. Apparently Mr. Phillips had in 1904 a machine capable of well-balanced flight, if he had made the rudders large enough, and provided a mechanism for rotating the slats at either wing end, so as to control the lateral poise, as proposed by the present writer in 1893, for practically that same flier (see page 229). Phillips’s aëroplane shows a distinct advance over its predecessors, even Wenham’s multiplane, because of the careful curving of the sustainers. Tatin’s flat wing machine had, indeed, shown a greater efficiency as a whole, but that was likely due to less proportionate body resistance. To Phillips we owe the introduction of superposed arched surfaces, now so commonly used in mechanical flight. Whether he was wise in using so many narrow wings, instead of a few broad ones, was a question to be answered by precise measurement.