660 Professor J. A. Fleming [June 4, 



for example, if we are propagating plane waves 1000 feet long over a 

 surface which by itself would reduce the wave amplitude to 0*367 of 

 its initial amplitude in 1000 kilometres, then, when we consider the 

 decrease by distance as well, we have to take account of the fact that 

 this last cause reduces the wave amplitude at 1000 kilometres to 

 0*001 of that which it is at 1 kilometre distance. I have represented 

 in the diagram some of Dr. Zenneck's curves. The dotted line shows 

 the decrease of amplitude by distance alone, and the firm lines that 

 due to distance and terrestrial absorption in various cases. We are 

 able to see from them the large effect due to travel over large distances 

 of very dry soil. Thus, for instance, if the absorption is such as to 

 cut down the amplitude in the ratio of 1 : 0*367 at 1000 kilometres, 

 then at a distance of 3000 kilometres the amplitude of a wave of 

 1000 feet in length would be cut down in the ratio of 3000 to 1 by 

 distance alone, but in the ratio of 60,000 to 1 by distance and terres- 

 trial absorption combined. 



An important matter is the question of the influence of wave- 

 length on this absorption. It can be shown from theory that an 

 increase of wave-length reduces the energy dissipation by the earth. 

 Thus in certain cases increasing the wave-length from lOoO to 10,000 

 feet increases the range of effective communication 100 times. The 

 absorption is also determined by the decrement of the wave-train 

 being greater the larger the decrement. 



One practical deduction to be made from this investigation is that 

 the reduction in wave amplitude which takes place when the wave 

 moves over very dry soil is as much due to small dielectric constant 

 of the material as to high resistivity. We see also that the wave 

 front is very far from being vertical when the waves travel overland, 

 and hence it is an advantage in that case for the receiving antenna to 

 slope away from the direction in which the waves are travelling or 

 from the radiant point. Lastly, it points to the advantage of a long 

 wave for overland working. Generally speaking, then, we find that 

 electric wave telegraphy is conducted with nmch greater ease over sea 

 than over dry land, the reason being that the dielectric constant is 

 large and the conductivity of sea water is sufficient to prevent much 

 penetration of the electric wave in the sea, and therefore there is not 

 much dissipation of its energy by absorption due to the surface over 

 which it travels. We have here an instance of economy in Nature. 

 Over sandy deserts, where we can, if need be, put up telegraph posts 

 and wires, radiotelegraphy has had some natural difficulties placed 

 in its way, but on sea, where connection between moving stations is 

 the important matter, and telegraph posts are impossible, special 

 facilities seem to have been afforded us for conducting it. 



The next point in connection with the antenna to be noticed is 

 the means adopted of setting up the oscillations in it. The universal 

 custom at present is to excite oscillations in a reservoir circuit con- 

 sisting of a condenser and an inductance by means of the spark or 



