Another great difficulty was to obtain a steady supply of power on the airplane to operate the transmitting-apparatus. It has been the practice to supply current on airplanes for wireless-telegraph apparatus by means of a small electric generator which is revolved by a little propeller. The propeller in turn is revolved by the rush of air as it is carried along by the plane. But the speed of the airplane varies considerably. At times, it may be traveling at only forty miles per hour, and at other times as high as one hundred and sixty miles per hour, so that the little generator is subjected to great variations of speed and consequent variations of voltage. This made it impossible to produce the steady oscillations that are required in wireless telephony. After considerable experiment, a generator was produced with two windings, one of which operated through a vacuum tube, somewhat like an audion, and to resist the increase of voltage produced by the other winding. Then another trouble developed. The sparks produced by the magneto in the airplane motor set up electro-magnetic waves which seriously affected the receiving-instrument. There was no way of getting rid of the magneto, but the wires leading from it to the engine were incased in metal tubes which were grounded at frequent intervals, and in that way the trouble was overcome to a large extent. The magnetos themselves were also incased in such a way that electro-magnetic waves would not be radiated from them. Instead of using trailing wires which were liable to become entangled in the propeller, the antenna was extended from the upper plane to the tail of the machine, and later it was found that by using two short trailing antennæ one from each tip of the wings, the very best results could be obtained. Still another development was to embed the antenna wires in the wings of the plane. It was considered necessary, if the apparatus was to be practicable, to be able to use it over a distance of two thousand yards, but in experiments conducted in October, 1917, a couple of airplanes were able to talk to each other when twenty-three miles apart, and conversations were carried on with the ground from a distance of forty-five miles. The conditions under which these distances were attained were unusual, and a distance of three miles was accepted as a standard for communication between airplanes. The apparatus weighed only fifty-eight pounds and it was connected with both the pilot and the observer so that they could carry on conversations with each other and could both hear the conversation with other airplanes or the ground. As a matter of fact, airplanes with standard apparatus are able to talk clearly to a distance of five miles and even to a distance of ten miles when conditions are favorable, and they can receive messages from the ground over almost any distance.