976 PROCEEDINGS OF SECTION H. 
Sundays, but the capacity of the cells is sufficient in case of need 
to light the beacons for three nights in succession without re- 
charging. 
For the beacon lighting, incandescent lamps of 16-candle 
power are employed, the voltage of the lamps ranging from 170 
to 100 volts with the distance of the beacons from the generating 
station. The lamps are not allowed to burn for more than 500 _ 
hours, being changed when that length of time is reached. 
To provide against the extinction of the light in any of the 
beacons through the failure of an incandescent lamp, a differential 
relay is placed in each of the lanterns, which automatically brings 
a spare lamp into circuit immediately the first lamp burns out. 
As in the leading lights used in the longer reaches of the river 
the power of the lamps is raised to about 70 candles by means 
of reflectors and lenses, it was necessary that the broken lamp 
should be thrown out of position, and the spare lamp brought 
exactly into the focal point. This is effected by mounting the 
two lamp-holders on a spindle at an angle of 90 degrees to one 
another. The extinction of the first lamp releases a trigger, 
freeing the spindle, which is revolved by means of a counterpoise 
weight, bringing the spare lamp into position, the cireuit for which 
1s completed by the fall of the trigger. A small oil dash-pot con- 
nected to the spindle gives it a slow motion, and prevents any 
jarring of the filament. 
For the purpose of checking the working of the lights a record- 
ing ammeter is placed in the main circuit, and from the resulting 
diagram it is possible to ascertain if any lamps have failed during 
the night, if the second lamp has been switched in, and, approxi- 
mately, from the deflection of the line, the position of the beacon 
in which this has taken place. The failures have been very few. 
A mercurial contact is attached to this ammeter, which rings 
an alarm-bell if the current falls below its proper point during the 
time the lights are burning. 
Tn conclusion, I crave pardon for the desultory character of the 
notes which I have submitted, but they will serve to show in 
some measure the trend of our departmental engineering in South 
Australia, and the details given may be useful for the sake of 
comparison with those of similar work elsewhere. That the 
possibility of such comparison may be more assured in the future, 
T beg to enter a plea for the unification of the methods of pre- 
paring estimates for engineering works and keeping the records 
thereof, especially in regard to the items of cost in the various 
parts of Australasia. 
In preparing estimates, or recording expenditure, I include all 
charges for carriage, interest on loan during construction, and 
reimbursements to other Departments, as well as the actual 
expenditure in money on material and labour. I have found that 
