266 



FKAGMENTS OF SCIENCE. 



the values of the sounds at the distances recorded woulc 

 not, in my opinion, show a greater advance with the in- 

 crease of material than that indicated by the foregoing 

 numbers. Subsequent experiments rendered still more 

 certain the effectiveness, as well as the economy, of the 

 smaller charges of gun-cotton. 



It is an obvious corollary from the foregoing experi- 

 ments that on our 'nesses' and promontories, where the 

 land is clasped on both sides for a considerable distance 

 by the sea where, therefore, the sound has to propagate 

 itself rearward as well as forward the use of the para- 

 bolic gun, or of the parabolic reflector, might be a 

 disadvantage rather than an advantage. Here gun- 

 cotton, exploded in the open, forms the most appropriate 

 source of sound. This remark is especially applicable 

 to such lightships as are intended to spread the sound 

 all round them as from central foci. As a signal in 

 rock lighthouses, where neither syren, steam -whistle, 

 nor gun could be mounted ; and as a handy fleet-signal, 

 dispensing with the lumber of special signal-guns, the 

 gun-cotton will prove invalua,ble. But in most of these 

 cases we have the drawback that local damage may be 

 done by the explosion. The lantern of the rock light- 

 house might suffer from concussion near at hand, and 

 though mechanical arrangements might be devised, 

 both in the case of the lighthouse and of the ship's deck, 

 to place the firing-point of the gun-cotton at a safe 

 distance, no such arrangement could compete, as regards 

 simplicity and effectiveness, with the expedient of a 

 gun-cotton rocket. Had such a means of signalling 

 existed at the Bishop's Rock lighthouse, the ill-fat 

 'Schiller' might have been warned of her approach to 

 danger ten, or it may be twenty, miles before she reached 

 the rock which wrecked her. Had the fleet possessed 

 such a signal, instead of the ubiquitous but ineffectual 



