The atmosphere, however, plays many strange pranks with the most powerful sound-producing instruments. To-day, for instance, a fog-signal may be heard at a distance of ten miles; to-morrow it will fail to be audible more than a mile away. This aberration of sound is extraordinary and constitutes one of the unsolved problems of science. Innumerable investigations have been made with the object of finding the cause of this erratic action, but no conclusive explanation has been forthcoming. Another strange trick is that, while a sound may be audible at distances of two and four miles during a fog, it fails to strike the ear at three miles. It is as if the sound struck the water at a range of two miles, bounded high into the air, and again fell upon the water at four miles, giving a second leap to hit the water again farther on, in much the same way as a thin flat stone, when thrown horizontally into the water, will hop, skip, and jump over the surface. This trick renders the task of the lighthouse engineer additionally exasperating and taxes his ingenuity to the utmost, as it appears to baffle completely any attempt towards its elimination. Recently another ingenious and novel system has been perfected by Messrs. D. and C. Stevenson. This is an acetylene gun which acts automatically. Hitherto an unattended fog-signal--except the bell-buoy tolled by the movement of the waves, which is far from satisfactory, or the whistling buoy, which is operated upon the same lines and is equally ineffective except at very short range--has found little favour. The objections to the bell and whistle buoys are the faintness of the sounds, which may be drowned by the noises produced on the ship herself; while, if the wind is blowing away from the vessel, she may pass within a few feet of the signal, yet outside its range. Thus it will be recognized that the fog-gun serves to fill a very important gap in connection with the warning of seafarers during thick weather. As is well known, even a small charge of acetylene, when fired, will produce a loud report, and this characteristic of the gas induced Messrs. Stevenson to apply it to a fog-signal. They have developed the automatic acetylene system of lighting to a very high degree around the coasts of Scotland, and there are now more than twenty lights of this class, mostly unattended, in operation, some of which have been established for many years. These lights have proved highly satisfactory. There has never been an accident, a freedom which is due to the fact that Moye’s system is used, wherein the possibilities of mishap are surmounted very effectively. Accordingly, the engineers saw no reason why a similar system should not be adapted to the emission of sound instead of light signals, or, if desired, of both simultaneously. Their experiments have been crowned with complete success, and, as the gun uses no more gas than would be consumed if a flashing light system were used, the cost of operation is very low.