514 TRANSACTIONS OF SECTION G. 
to 0-005 joule causing ignition. The influence of gas in forwarding coal-dust 
explosions begins to be felt when 4 per cent. of gas is present. At 2 per 
cent. full ignitions are obtained at every trial. Coal-dust alone can be ignited 
by both continuous-current or alternating-current break-flashes, the former 
requiring 3°5 to 6 amperes at 480 volts in non-inductive circuits, the latter 
14 amperes at 40 periods and on a power factor of 0°8. Continuous-current 
faults on the negative cable develop rapidly in the presence of moisture and 
the cable is disintegrated. Alternating-current faults are self-healing, and a 
mechanical fault does not increase electrically on an alternating-current cable. 
Armouring is necessary under modern power conditions; lighting and sig- 
nalling circuits must be equally well protected to prevent open sparking. The 
limits of safety are electrically low, but the risks of ignition are even now no 
greater than those attending the use of flame safety lamps, and they can be 
entirely prevented. 
5. The Balsillie System of Wireless Telegraphy as employed in the 
Radio-Telegraph Stations of the Commonwealth of Australia.+ 
By J. G. Bausiuuig. 
6. The Capacity of Radio-Telegraphic Aerials.? 
By Professor G. W. O. Hows, D.Sc. 
The capacity considered in this paper is the actual static capacity, and not the 
equivalent capacity of the antenna considered as part of an oscillatory circuit. 
The accurate calculation of the capacity of a multiple-wire horizontal aerial 
with its leading-down wires would be a difficult mathematical problem, quite 
unwarrantel by the practical requirements of radio-telegraphy. 
When raised to a potential above or below that of the earth, the charge is 
distributed over the antenna in a way which is not easy to calculate, but which 
must be such that all parts of the antenna are at the same potential. If the 
antenna were made up of a great number of short pieces, placed end to end, 
but insulated from each other, it would be possible to distribute the charge 
uniformly, but the potential would then vary from point to point in a way which 
is easily calculated. If now we assume that all the separate pieces of wire are 
connected, electricity will flow from points of high to points of low potential 
until the potential is everywhere the same. The assumption made in developing 
the various formule is that this final uniform potential is equal to the average 
value of the potential when the charge was uniformly distributed. ‘This is only 
approximately correct, but the accuracy is more than sufticient for all practical 
purposes. ‘This method has been applied to antenne of all the types usually 
employed, and formule have been established for each type. A large number of 
numerical examples have been worked out, and the results are given in tables and 
curves, so that the capacity of any antenna can be read off directly from its 
dimensions. ‘The corrections due to the leading down wires and to the proximity 
of the earth are fully considered, and examples given showing the application of 
the formule to antennz of any type. 
Formule for the calculation of the capacity of antenne have recently been 
published by Pedersen (‘Jahrbuch der Drahtlosen Telegraphie,’ vii. 4, p. 434) 
and Louis Cohen (Hlectrician, February 14 and 21, 1913). When applied 
to the experimental results quoted by Cohen, the formule developed in this 
paper give closer agreement than do the formule given by Cohen, although some 
of his results can hardly be reconciled with the given data. 
7. Irrigation Dams and Hydro-Electric Power.* 
By E. Krezurn Scorr. 
1 Published in the Hlectrician, Ixxiv., p. 70. 
2 Published in the Hlectrician, \xxili., p. 829. 
° Published in Llectrical Review, \xxv., p 317. 
