There are various methods of consummating this end. One method is to haul in the balloon and to peg it down on all sides, completing the anchorage by the attachment of bags filled with earth to the network. While this process is satisfactory in calm weather, it is impracticable in heavy winds, which are likely to spring up suddenly. Consequently a second method is practised. This is to dig a pit into the ground of sufficient size to receive the balloon. When the latter is hauled in it is lowered into this pit and there pegged down and anchored. Thus it is perfectly safe during the roughest weather, as none of its bulk is exposed above the ground level. Furthermore it is not a conspicuous object for the concentration of hostile fire. In some instances, and where the military department is possessed of an elaborate equipment such as characterises the German army, when reconnaissance is completed and the balloon is to be removed to another point, the gas is pumped back into the cylinders for further use. Such an economical proceeding is pretty and well adapted to manoeuvres, but it is scarcely feasible in actual warfare, for the simple reason that the pumping takes time. Consequently the general procedure, when the balloon has completed its work, is to permit the gas to escape into the air in the usual manner, and to draw a fresh supply of gas from further cylinders when the occasion arises for re-inflation. Although the familiar spherical balloon has proved perfectly adequate for reconnoitring in the British and French armies, the German authorities maintained that it was not satisfactory in anything but calm weather. Accordingly scientific initiative was stimulated with a view to the evolution of a superior vessel. These endeavours culminated in the Parseval-Siegsfeld captive balloon, which has a quaint appearance. It has the form of a bulky cylinder with hemispherical extremities. At one end of the balloon there is a surrounding outer bag, reminiscent of a cancerous growth. The lower end of this is open. This attachment serves the purpose of a ballonet. The wind blowing against the opening, which faces it, charges the ballonet with air. This action, it is claimed, serves to steady the main vessel, somewhat in the manner of the tail of a kite, thereby enabling observations to be made as easily and correctly in rough as in calm weather. The appearance of the balloon while aloft is certainly curious. It appears to be rearing up on end, as if the extremity saddled with the ballonet were weighted.