284 THE POPULAR SCIENCE MONTHLY. 



The day was cloudy, with occasional showers of drizzling rain ; the 

 wind about northwest by north all day ; at times squally, rising to a 

 force of 6 and 7 and sometimes dropping to a force of 2 or 3. The 

 station at Leigh excepted, all these places were to leeward of Shoe- 

 buryness. At four other stations to leeward, varying in distance from 

 15^ to 24|- miles, nothing was heard, while at eleven stations to wind- 

 ward, varying from 8 to 26 miles, the sounds were also inaudible. It 

 was found, indeed, that the sounds proceeding against the wind did not 

 penetrate much beyond 3 miles. 



On the following day, viz., the 18th of October, we proceeded to Dun- 

 geness with the view of making a series of strict comparative experi- 

 ments with gun-cotton and cotton-powder. Rockets containing 8 ounces, 

 4 ounces, and 2 ounces of gun-cotton had been prepared at the Royal 

 Arsenal ; while others, containing a similar quantity of cotton-powder, 

 had been supplied by the Cotton-Powder Company at Faversham. 

 With these were compared the ordinary 18-pounder gun, which hap- 

 pened to be mounted at Dungeness, firing the usual charge of 3 pounds 

 of powder. 



From these experiments it appeared that the gun-cotton and cotton- 

 powder were practically equal as producers of sound. 



The effectiveness of small charges was also illustrated in a very 

 striking manner, only a single unit separating the numerical value of 

 the 8-ounce rocket from that of the 2-ounce rocket. The former was 

 recorded as 6.9 and the latter as 5.9, the value of the 4-ounce charge 

 being intermediate between them. These results were recorded by a 

 number of very practised observers on board the Galatea. They were 

 completely borne out by the observations of the coast-guard, who marked 

 the value of the 8-ounce rocket 6.1, and that of the 2-ounce rocket 5.2. 

 The 18-pounder gun fell far behind all the rockets, a result probably to 

 be in part ascribed to the imperfection of the powder. The performance 

 of the siren was, on the whole, less satisfactory than that of the rockets. 

 The instrument was worked, not by steam of 70 pounds pressure, as at the 

 South Foreland, but by compressed air, beginning with 40 pounds and 

 ending with 30 pounds pressure. The trumpet was pointed to wind- 

 ward, and in the axis of the instrument the sound was about as effective 

 as that of the 8-ounce rocket. But in a direction at right angles to the 

 axis, and still more in the rear of this direction, the siren fell very sensi- 

 bly behind even the 2-ounce rocket. 



These are the principal comparative trials made between the gun- 

 cotton rocket and other-fog signals ; but they are not the only ones. On 

 the 2d of August, 1877, for example, experiments were made at Lun- 

 dy Island, with the following results : At 2 miles distant from the firing- 

 point, with land intervening, the 18-pounder, firing a 3-pound charge, 

 was quite unheard. Both the 4-ounce rocket and the 8-ounce rocket, 

 however, reached an elevation which commanded the acoustic shadow, 

 and yielded loud reports. When both were in view, the rockets were 



