The other chief French experimenters at this period were the Voisin Freres, whose first two machines--identical in form--were sold to Delagrange and H. Farman, which has sometimes caused confusion, the two purchasers being credited with the design they bought. The Voisins, like the Wrights, based their designs largely on the experimental work of Lilienthal, Langley, Chanute, and others, though they also carried out tests on the lifting properties of aerofoils in a wind tunnel of their own. Their first machines, like those of Santos-Dumont, showed the effects of experimenting with box-kites, some of which they had built for M. Ernest Archdeacon in 1904. In their case the machine, which was again a biplane, had, like both the others previously mentioned, an elevator in front--though in this case of monoplane form--and, as in the Wright, a rudder was fitted in rear of the main planes. The Voisins, however, fitted a fixed biplane horizontal 'tail'--in an effort to obtain a measure of automatic longitudinal stability--between the two surfaces of which the single rudder worked. For lateral stability they depended entirely on end curtains between the upper and lower surfaces of both the main planes and biplane tail surfaces. They, like Santos-Dumont, fitted a wheeled undercarriage, so that the machine was self-contained. The Voisin machine, then, was intended to be automatically stable in both senses; whereas the Wrights deliberately produced a machine which was entirely dependent upon the pilot's skill for its stability. The dimensions of the Voisin may be given for comparative purposes, and were as follows: Span 33 feet with a chord (width from back to front) of main planes of 6 1/2 feet, giving a total area of 430 square feet. The 50 horse-power Antoinette engine, which was enclosed in the body (or 'nacelle ') in the front of which the pilot sat, drove a propeller behind, revolving between the outriggers carrying the tail. The total weight, including Farman as pilot, is given as 1,540 lbs., so that the machine was much heavier than either of the others; the weight per horse-power being midway between the Santos-Dumont and the Wright at 31 lbs. per square foot, while the wing loading was considerably greater than either at 3 1/2 lbs. per square foot. The Voisin machine was experimented with by Farman and Delagrange from about June 1907 onwards, and was in the subsequent years developed by Farman; and right up to the commencement of the War upheld the principles of the box-kite method of construction for training purposes. The chief modification of the original design was the addition of flaps (or ailerons) at the rear extremities of the main planes to give lateral control, in a manner analogous to the wing-warping method invented by the Wrights, as a result of which the end curtains between the planes were abolished. An additional elevator was fitted at the rear of the fixed biplane tail, which eventually led to the discarding of the front