WATERWAYS OF CULVERTS AND BKIDGES. 557 



character of the surface of remote parts of the basin. The local 

 high flood mark is sought, and either witli or without the assistance 

 of that venerable but very unscientilic personage, " the oldest 

 inhabitant," is supposed to be found. It is then shown upon the 

 longitudinal section, and the area of opening is arbitrarily fixed by 

 some one at the head-office, who very often has no real knowledge 

 of the country and the streams, a large depression being 

 favoured with a compai'advely large bridge, while a small gap is 

 pi'ovided with an insignificant culvert, no account being taken of 

 the area drained, of the rainfall statistics, or the slope and general 

 character of the ground. A fter a time heavy rains occur, culverts 

 and bridges fail, traffic is impeded, perhaps altogether stopped, 

 and costly re-construction takes place — in some instances followed 

 by still further failures and yet more costly re-construction at a 

 still later date Say not that this picture is exaggerated. Did I 

 wish I could name the date and locality of its occurrence. 



This too common mode of treatment is a most fallacious one. 

 In the first place, the recollection of the "oldest inhabitant" is 

 often of the vaguest kind. I have known ajiparently intelligent 

 pei'sons point out flood levels that other and indubitable evidence 

 proved to be several feet in error within two years of the occurrence 

 of a disastrous flood. They almost always exaggerate the height 

 of the water. On the other hand, flood-marks consisting of 

 rubbish caught in branches of trees, or lines of debris along the 

 .slope, may be, and often are too low. This is due either to the fact 

 that the rubbish did not reach the locality where it was stranded 

 until the flood had partly subsided, or that it truly marks the 

 highest level of a recent but comparatively small flood, the traces 

 of the great flood of several years earlier date having long ago 

 vanished. Hence flood-marks and evidence of residents are 

 frecjuently more or less misleading. Then again, in thinly-peopled 

 districts it often happens that no personal testimony is available, 

 and at the same time no distinct lines of debris ax"e to be found, 

 so that the surveyor is utterly without local data to go upon. 



Further, the appearance of the longitudinal .section of the road 

 or railway is often deceptive. What bears the aspect of almost a 

 river may merely be a local depression, a pond of still water that 

 may be embanked across with impunity, while a comparatively 

 shallow dip or narrow cleft may really be the sole outlet of the 

 drainage of a large district, to curtail which in the slightest 

 degree would be a most dangerous proceeding. 



Having frequently had to deal with questions of waterway, I 

 have made it my business to collect the best available information 

 botli in the way of rules laid down by authors, and in the way of 

 local Australian experience, and I now propose to submit the 



