42 Quantity of Water Consumed in Irrigation. 



should be considered as merely tentative — to be replaced 

 by a better scale when fuller information is available. 



31. However vague and imperfect the results given in 

 this paper may be considered, it should not be forgotten that 

 the same points have been hitherto not less vaguely dealt 

 with (see par. 3), and that, moreover, irrigation is not 

 singular in this respect. In many engineering calculations 

 there is often very great latitude allowed in the shape of 

 co-efficients, the value of which depends on the judgment of 

 the one using them ; as, for instance, in calculating discharges 

 of an irregular channel, of a channel when partially 

 obstructed and, generally, whenever the conditions are com- 

 plicated, as in practice they often are. Many other points, 

 especially in hydraulic engineering, might be mentioned, 

 such as the proportion of rainfall flowing off the surface of 

 the ground for . the supply of a reservoir, the volume of 

 sewage it is necessary to provide for in any case, and others 

 which will readily occur to an engineer. 



32. Further investigation being very desirable, or rather es- 

 sentially necessary, a few remarks on the point may be useful. 

 When measuring the quantity of water used in irrigation, in 

 order that results obtained in different localities may be 

 compared, the author would suggest that the following par- 

 ticulars be noted : — Yolume of water supplied, length of 

 channel from the point of gauging the discharge to the field, 

 mean wetted perimeter of channel, whether the channel is old 

 or newly formed, and whether puddled in any way or not, 

 average size of the beds in the fields, or the approximate 

 number per acre, how long one bed takes to fill, how many 

 previous waterings have been given to the same crop, at 

 what intervals and how long since the last, whether or not 

 the watercourse has been used just before for another field 

 or otherwise, if the field has been hoed or the surface of the 

 ground disturbed since the last watering, height of the 

 crop or the extent to which it shelters the ground, nature 

 of the soil, state of the weather at and just before the time 

 of watering, date of measurement, and such other informa- 

 tion as the observer may deem likely to affect the result. 

 For want of full particulars, results hitherto obtained in dif- 

 ferent places are often not comparable. If, moreover, the 

 observer would endeavour to apportion the total con- 

 sumption between the fields and the watercourses in each 

 case, somewhat after the example given above (see par. 27), 

 it is probable that correct information would be obtained 



