TtADIOTELEGRAPHY MARCONI. 



121 



almost completely the spark gap and its consequent resistance, which, 

 as is well known, is the principal cause of the damping or decay of 

 the waves in the usual transmitting circuit. » 



The apparatus shown in figure 6 consists of a metal disk a, having 

 copper studs firmly fixed at regular intervals in its periphery and 

 placed transversely to its plane. This disk is caused to rotate very 

 rapidly between two other disks, b, by means of a rapidly revolving 

 electric motor or steam turbine. These side disks are also made to 

 slowly turn round in a plane at right angles to that of the middle 

 disk. The connections are as illustrated in the figure. The studs 

 are of such length as to just touch the side disks in passing, and 

 thereby bridge 



RL5DNflNCC CURVL 

 orCLIFDEN 

 PRIMARY CIRCUIT 



TOTAL RESISTANCE __^j22 

 INCLUOING, SPARK, ~ 



3 S s g ft « *S jS j» S gj Wave Lcn 



us4V£aG>Sg£S3rS ik| FEET 



O5CILLATI0N5 CORRESPONDING TO ABOVE RE50NANCE CURVE 



the gap between ^ £ £) 



the latter. 



With the fre- 

 quency employed 

 at Clifden, name- 

 ly, 45,000, when 

 a potential of 

 15,000 volts is 

 used on the con- 

 denser, the. spark 

 gap is practically 

 closed during the 

 time in which one 

 complete oscilla- 

 tion only is taking 

 place, when the 

 periphe ric al 

 speed of the disk 

 is about 600 feet a 

 second. The re- 

 sult is that the 



primary circuit can continue oscillating without material loss by 

 resistance in the spark gap. Of course the number of oscillations 

 which can take place is governed by the breadth or thickness of the 

 side disks, the primary circuit being abruptly opened as soon as the 

 studs attached to the middle disk leave the side disks. 



This sudden opening of the primary circuit tends to immediately 

 quench any oscillations which may still persist in the condenser 

 circuit; and this fact carries with it a further and not inconsiderable 

 advantage, for if the coupling of the condenser circuit to the aerial 

 is of a suitable value the energy of the primary will have practically 

 all passed to the aerial circuit during the period of time in which the 







jw 



^2. 



"vf. 



& 



5ERIE5. or WAVE TRAINS RADJATEP 

 5PARK FREQUENCY 500 



Fig. 5. 



_4£5tcs. 



*om CLIFDEN AERIAL 

 PER 5K. 



