"She's holding, I fancy," said Fuller referring to the suspected float. "Anyhow we've kept afloat so far and there's no reason why we shouldn't do so until I tackle this most refractory motor."
Making cautious use of the flash lamp the pilot minutely examined the complicated mechanism. It was not long before the mischief was discovered. Not only was the petrol-tank completely perforated by three shrapnel bullets, but the pipe leading from it to the carburetter had been cut clean through. That accounted for the engine running for some seconds before coming to a stop. Until the last of the petrol in the carburetter had been drawn into the cylinders firing was still taking place.
Further examination revealed the fact that, the motor was otherwise undamaged, although, judging by the holes in the fuselage and through the planes, it seemed wonderful that pilot and observer had escaped being hit.
"Can I bear a hand?" enquired Kirkwood.
"No, thanks," was the reply. "Close enough quarters as it is. We should only be tumbling over one another."
By the aid of a piece of flexible tubing lined with indiarubber the broken portions of the petrol pipe were temporarily reunited. The next step was to plug the holes in the tank. This task was performed by means of a metal instrument consisting of a metal rod of about a third of an inch in diameter and four inches in length. Two thirds of the length was threaded and fitted with a "butterfly" nut in front of which was a cylindrical plug of guttapercha faced with indiarubber. At the other extremity was a swivelled cross-bar of about an inch in length and so arranged that it could lie in a straight line with the rod.
This end Fuller inserted in one of the perforations in the side of the tank. Then, giving the rod half a turn, he allowed the swivel bar to fall into a position at right angles to the rod. It was then impossible to withdraw the latter owing to the cross-piece engaging on the inner side of the tank.
The flight-lieutenant's next move was to screw the pliable plug hard against the perforated metal by means of the "butterfly" nut, and by so doing hole No.1 was repaired—the first of six. While Fuller was engaged upon the work of making good defects Kirkwood, his mind still uneasy on the subject of the float, lowered himself over the side.
Gaining the upper side of the float he felt along it with his hand. As he did so a wave swept the frail buoyant structure.
"By Jove!" he exclaimed. "This is a treat. The water is quite warm."