Quantity of Water Consumed in Irrigation. 41 



by the various channels, the approximate depth of soil and 

 the nature of the subsoil, must be ascertained. Then, take 

 the mean wetted perimeter of the canal and each dis- 

 tributary and multiply it by the respective length of each 

 and by the number of days water would be flowing during 

 the irrigating season. Allow a loss according to the scale 

 given in Appendix A (cols. 2 and 3) over the whole wetted 

 surface of bed and slopes of all channels.* If the soil or 

 subsoil in which the canal is carried is an open or loose 

 gravel or is much fissured and of considerable depth, a 

 much larger loss than given by the scale may be expected ; 

 such places should be avoided if possible. In a narrow 

 valley, or where an impervious substratum lies a little 

 below the canal bed, or at a greater depth but rising on 

 both sides so as to confine the subterranean current, or 

 where the subsoil is less pervious than the surface soil, the 

 loss would probably be less. Where, however, canals of any 

 size are likely to be made, the country would be more or 

 less open and the underground current practically un- 

 restricted. After a few years the loss might be expected to 

 decrease, owing to the pores of the subsoil becoming 

 clogged. 



VI — Concluding Remakks. 



30. The author has for several years watched carefully 

 the working of various irrigation projects, and made inquiry 

 into the causes of failure in certain cases. As a contribution 

 to a subject on which but little is accurately known, loose 

 generalisations being usually accepted, he thinks it right to 

 make public the results of his observations, imperfect as they 

 are ; others can then make use of them or not as they please. 

 Much difference of opinion may be expressed regarding the 

 real value of the data accepted by the author, and the con- 

 clusions he has arrived at — (he will be glad to see these fully 

 criticised), but he would urge on those having better know- 

 ledge of the subject, that the results of their observations 

 should at the same time be given in a form suitable for the 

 object of this paper — to serve as a guide for engineers in 

 other countries than their own, epsecially in countries like 

 Australia, where irrigation is in its infancy. Appendix A 



* Where experiments on the absorbing power of the soil can be made in 

 the manner described a little further on (see par. 33), it would perhaps be 

 more satisfactory to base calculations of loss on the results so obtained,, 



E 



