While these indefatigable efforts were in progress, ingenious attempts were made to press Nature herself into operation. As is well known, there are many “blowing-holes” distributed throughout the world, where the water by erosion has produced a long, narrow cavern in the base of a rock, with a constricted outlet into the outer air. The waves, rushing into the cave, compress the air within, which, in its escape at high velocity through the small vent, produces a bellowing sound. It was this curious phenomenon which gave the Wolf Rock its name. General Hartmann Bache, of the United States Engineers, attempted in 1858 to make use of a blowing-hole on one of the Farallon Isles, lying forty miles off the entrance to San Francisco Bay. A chimney was built with bricks above the orifice, through which the air compressed by the waves below made its escape, and on top of this shaft a locomotive whistle was placed. The first effort was a dead failure, because the force of the rush of air was so great that it carried away the chimney; but in the second attempt success was achieved, and an excellent automatic whistle blared out night and day almost continuously and was audible for some distance out to sea. The only drawback was that in foggy weather, when the most intense sound was required, the signal was dumb owing to the smoothness of the water. This novel signal was maintained for some time and then was superseded by a powerful siren. One of the most interesting fog-signalling installations in service is that on the bald formidable hump of rock lying in the estuary of the Clyde, known as Ailsa Craig. For years this rock constituted a terrible menace to the crowded shipping of this important marine thoroughfare, and its victims were numerous. While the Commissioners of Northern Lighthouses mitigated its terrors as far as possible by the provision of a powerful light, they recognized the fact that a visual warning did not meet the situation completely. But the installation of a fog-signal was a somewhat peculiar problem, owing to the configuration of the rock. A single station would not meet requirements, because it was necessary to throw the warning from both sides of the obstruction. The provision of two sound-stations would have been an expensive matter, even if it had been feasible, which it was not, owing to the precipitous nature of the cliffs.