730 Experiments on the communication of QSept. 



of the Hooghly being interposed) and proceeded to the north extremity 

 of a nullah 3,000 feet in length; from the south end of the nullah a wire 

 returned to the library. Thus we had altogether eleven miles of me- 

 tallic and 13,256 feet of water circuit, the latter in two interruptions. 

 The signal still passed as intelligibly and strongly as before. A litho- 

 graphic plan is annexed in illustration of these details. 





§ 7- — General remarks on the applicability of the preceding facts. 



I reserve for another occasion the description of several experiments 

 which these facts led to, regarding the possibility of dispensing alto- 

 gether with metallic conductors; and I now proceed to shew some 

 practical circumstances, expenses, and other details regarding the ap- 

 plication of these facts to the actual accomplishment of telegraphic 

 correspondence. 



To effect a perfect system of telegraphic communication for a dis- 

 tance, say of 500 miles, two wires are at most required ; where a river 

 passes between the termini, but one wire is necessary. If the water 

 communication be the sea, the distance for which one wire will suffice 

 will be at least quadrupled. Insulation of one wire is necessary. The 

 wire may touch dry substances of any kind, but it must not come in 

 contact with water or damp earth communicating with the second 

 wire or with the river course. 



Insulation according to my experiments is best accomplished by 

 enclosing the wire (previously pitched) in a split ratan, and then paying 

 the ratan round with tarred yarn — or the wire may, as in some ex- 

 periments recently made by Colonel Pasley in Chatham, be sur- 

 rounded by strands of tarred rope, and this by pitched yarn. 



An insulated rope of this kind may be spread along a wet field, 

 nay, even led through a river, and will still conduct without any 

 appreciable loss the electrical signals above described. 



In establishing a communication it would be advisable to bury 

 the wire about two feet below the surface of the ground, in a narrow 

 trench well rammed with pounded brick and mortar. At every ten 

 miles the wire should rise through the ground in a masonry pillar, to 

 allow of verification or of the discovery of the situation of accident. In 

 India the Police Thannah houses might be conveniently used for this 

 purpose. 



The expense of copper wire per mile would be 272 rupees ; of 

 insulation 20 ; of trenching and masonry I can form no accurate 

 estimate. 



