556 PROCEEDINGS OF SECTION J. 



The Yarra at Melbourne flooded enormously in 1863, but since 

 then has been comparatively mild in its behaviour. The Goulburn, 

 Campaspe, and other streams on the northern slope of Victoria, 

 had their historic flood in 1870, and then subsided into comparative 

 harmlessness. From this and other instances the conclusion is 

 irresistible, that an experience varying from ten years in the case of 

 small, to thirty years in the case of moderately large streams, is the 

 very least that can justify an engineer in affirming, on the ground 

 of local experience purely, the sufficiency of a given waterway. 



This long experience of the behaviour of streams is generally 

 available in old and thickly-peopled countries. Floods in time 

 past have been watched, gauged, and are well remembered by those 

 who have suffered inconvenience and loss. Existing roads and 

 other works of old standing afford valuable precedents, so that the 

 engineer cannot well go wildly astray without the inconsistency of 

 his practice with what has gone before being noticed by the public. 

 Hence I think it is that writers such as Professor Rankine and 

 others, who are so clear and definite on other points, pay so little 

 attention to this. 



In new and thinly-peopled countries like the Australian colonies, 

 this local information is often unattainable. The engineer has no 

 local experience to guide him or precedents to follow. Hence the 

 necessity for specially careful consideration. Further, it is to be 

 mentioned, that in connection with railway work at anyrate, the 

 small and apparently unimportant culverts need even more care 

 and attention than large and costly bridges. A large bridge is 

 almost invariably close to a centre of population. During flood 

 time it is watched by many people ; further, the floods of large 

 streams rise gradually, and if a disaster occurs there is abundant 

 warning, so that while an error in such a case may result in the 

 destruction of a bridge, it is very unlikely to cause the wreck of a 

 train. A small culvert, on the other hand, is often found in a 

 remote and uninhabited place, the floods on its stream rise with 

 extreme suddenness and fall with equal celerity, perhaps unobserved 

 by any person. The bank may be washed away and a train plunge 

 into the chasm as has already happened in one most notable case, 

 and has nearly happened in another, without the slightest warning, 

 the attention of the officials being all the while directed to some 

 other menaced spot, and so diverted from the real point of greatest 

 danger. 



A method that has been too commonly adopted in fixing the 

 waterways is the following : — The surveyor fixes his attention 

 entirely upon that point of the stream whei-e the road or railway 

 crosses, but does not trouble to extend his enquiries up stream. 

 He knows neither the area drained, the length of the water- 

 courses, its average slope, the rainfall to be expected, nor the 



