'If it is calm, one must run a few steps down the hill, holding the machine as far back on oneself as possible, when the air will gradually support one, and one slides off the hill into the air. If there is any wind, one should face it at starting; to try to start with a side wind is most unpleasant. It is possible after a great deal of practice to turn in the air, and fairly quickly. This is accomplished by throwing one's weight to one side, and thus lowering the machine on that side towards which one wants to turn. Birds do the same thing--crows and gulls show it very clearly. Last year Lilienthal chiefly experimented with double-surfaced machines. These were very much like the old machines with awnings spread above them. 'The object of making these double-surfaced machines was to get more surface without increasing the length and width of the machine. This, of course, it does, but I personally object to any machine in which the wing surface is high above the weight. I consider that it makes the machine very difficult to handle in bad weather, as a puff of wind striking the surface, high above one, has a great tendency to heel the machine over. 'Herr Lilienthal kindly allowed me to sail down his hill in one of these double-surfaced machines last June. With the great facility afforded by his conical hill the machine was handy enough; but I am afraid I should not be able to manage one at all in the squally districts I have had to practice in over here. 'Herr Lilienthal came to grief through deserting his old method of balancing. In order to control his tipping movements more rapidly he attached a line from his horizontal rudder to his head, so that when he moved his head forward it would lift the rudder and tip the machine up in front, and vice versa. He was practicing this on some natural hills outside Berlin, and he apparently got muddled with the two motions, and, in trying to regain speed after he had, through a lull in the wind, come to rest in the air, let the machine get too far down in front, came down head first and was killed.' Then in another passage Pilcher enunciates what is the true value of such experiments as Lilienthal--and, subsequently, he himself--made: 'The object of experimenting with soaring machines,' he says, 'is to enable one to have practice in starting and alighting and controlling a machine in the air. They cannot possibly float horizontally in the air for any length of time, but to keep going must necessarily lose in elevation. They are excellent schooling machines, and that is all they are meant to be, until power, in the shape of an engine working a screw propeller, or an engine working wings to drive the machine forward, is added; then a person who is used to soaring down a hill with a simple soaring machine will be able to fly with comparative safety. One can best compare them to bicycles having no cranks, but on which one could learn to balance by coming down an incline.'