34 Quantity of Water Consumed in Irrigation. 



tributaries. Further calculations by the author, to ascer- 

 tain the loss in various soils (see App. D), show that a depth 

 of 1 \ feet would probably be lost daily in very sandy soil ; 

 in which case the loss in one part of the canal (the Northern 

 division) during this season would be sufficient to fill a 

 trench nearly 170 feet in width by over 200 feet in depth, 

 for the whole length of the division (about 50 milesj. In 

 the first fifteen miles or so, the loss might be double this, 

 or even more (see App. D). 



17. The quantity of water lost, when expressed in this 

 form, may appear enormous ; but further considerations will 

 perhaps convince those acquainted with the facts to be pre- 

 sently noticed, that it is not more than there is good reason 

 for believing actually takes place. The above quantity, 

 large as it may appear, is, however, small compared with the 

 average volume of water carried by the canal ; and special 

 measures to prevent the loss are not worth undertaking. 

 The average loss in the Northern division of the canal was, 

 in fact, only 15 - 58 cubic feet per second per lineal mile, out 

 of a mean discharge in the season of 4447 cubic feet per second 

 — about one-third ("35) per cent. only. In this form the loss 

 appears insignificant. Where, however, the canal is smaller, 

 and particularly in the distributaries, the proportion lost is 

 much greater. In a small watercourse the loss in a mile 

 may be one-fourth of the discharge at its head (see instance 

 given in next paragraph) ; here the advantage of puddle in 

 sandy soil is apparent. 



18. A Ganges canal officer (Mr. Beresford) some time since 

 gave instances of the loss of water in portions of his (the 

 Anupshahr) division,* which the author, from his recollection 

 of it, would consider not uncommon in that division. Mr. 

 Beresford mentioned a loss of 125 cubic feet per second in 

 the first mile of a distributary, having a discharge at the 

 head of 50 cubic feet per second ; allowing a wetted peri- 

 meter of 16 feet, the depth percolating in 24 hours would be 

 1*28 feet. In the case of a small watercourse having a head 

 discharge of one cubic foot per second, the loss was "03 

 cubic foot in a furlong (*24 cubic foot in a mile) ; with a 

 wetted perimeter of 4 feet, the depth lost would be *98 foot 

 per diem ; with 3 feet wetted perimeter, the loss would be 

 1*31 feet in depth. Mr. Login, who was some years ago in 



*Professional Papers on Indian Engineering, Vol. V., new series (Eoorkee, 

 1876), page 416. 



