2t Pi'oceadi i)(j-^ of the Royal Society of Victoria. 



amount of eri-ov iu the section of a river taken in a strong 

 current, and with probably a varying level of water. 

 Whether any correct proportion can be arrived at between 

 observations' taken in this way, and tliose taken with floats 

 in the ordinary manner, remains to be seen ; but as really 

 rolial)lo results cannot be obtained without a very large 

 nunil.)er of scientiticalh^ conducted experiments, it will 

 lirobaltl}' be a long time before any very great progress is 

 made in the gauging of rivers and streams, and as the cost 

 of observ^ations taken in the way above named is very 

 considerably greater than those taken in the ordinary 

 manner, it is to be feared that this method will not be very 

 generally practised. It would be very desirable if some 

 arrangement could be made by which the various scientitic 

 bodies, eno-ineers, and o-overnments would communicate to 

 each other any observations that were made by them, and 

 the results that each deduced from them. Meanwhile the 

 ordinary observations should be made systematically, and by 

 tlie very best operators obtainable, but even then it will be 

 necessaiy to be guided a good deal by the opinions of those 

 who from experience are able to estimate approximately the 

 amount of rini off in proportion to the rainfall in different 

 diotiicts and under different conditions. 



The next thing is to select the sites for reservoirs in 

 comiection with tlie land to be iirigated. For this ])urpose 

 men of special experience are required, as they must be 

 able to choose sites favourable for the construction of 

 reservoirs, judge approximately whether the rainfall 

 over the catchment area would be sufficient to till it, 

 whether if tilled it would provide foi' the irrigation 

 of all the land under its command, and to do this they 

 must be able to judge what land would be irrigable 

 from it. The site for a reservoir should be selected to 

 connnancl an area proportioned to its capacity, and the 

 probable length of drought must also be taken into 

 consideration. The de})th of the water sliould be great, so 

 as to reduce the proportion of loss b}^ evaporation. The 

 dam or bund to impound the water should be as short 

 as possible, to save exj)ense. The foundation inust be 

 carefully examined to see that it is suitable, and can be 

 made water-tight. Notice must also be taken of what 

 material is available for its construction, so as to determine 

 whether it would be most advantageously constructed of 

 earth or masonry. There must be a satisfactory site for 



