LIGHT RAILWAYS FOR NEW SOUTH WALES. 223 
break of gauge would involve, very great consideration should be 
given to the question. 
Mr. H.W. Parkinson—considered that as the function of a rail- 
way was to remove goods and passengers from one place to another, 
the line that enabled that to be done with the least expenditure 
of time, labour, and money, was the true light railway. To con- 
fine attention solely to first cost, and neglect working expenses, 
‘would therefore be wrong. In considering the possibility of 
reducing the first cost of lines, it would be well to divide that 
cost into two parts: (A), a known and for any given line constant 
cost, and (X), an unknown and variable amount. (A) consisting 
of the cost of forming, ballast, sleepers, rails, etc., which for 
different designs might vary from £2,000 to £4,000 a mile, or a 
total variation of £2,000. (X) consisting of earthworks and 
waterways might vary from £1,000 or less to say £17,000, or a 
possible variation of £16,000. There was, therefore, generally 
far greater scope for reduction in (X) than in (A), although 
attention was more often directed to the latter. A so-called 
reduction, however, would be no reduction at all if it entailed an 
equal or greater expenditure in working expenses. The suggested 
reduction in the depth of the ballast to three inches seemed to 
him of questionable utility—certainly on portions of many roads 
it would prove very inadequate. If the line were designed for a 
definite and constant axle load, then since the stiffness increased 
inversely as the cube of the span but only directly as the square 
of the weight of the rail per yard, evidently economy would be 
consulted by using as many sleepers as possible. This number, 
to leave proper room for tamping, would be about two thousand 
five hundred per mile. He was glad to note that the author was 
in favour of sharp curves as a means of reducing capital expendi- 
ture in construction. There could hardly be any doubt that the 
adoption of five chain curves on branch lines would very materially 
reduce their first cost without any corresponding increase in work- 
ing expenses. The mechanical difficulty of running round such 
curves with locomotives had been practically overcome, the greatest 
