Evaporation Results in New South Wales and South Africa. 



11 



into the natural water-bearing strata. Doubtless the like loss took 

 place all over the then surface of the dam. On the other hand, the 

 water table thus became permanently raised, and it is no longer 

 possible for so large a proportionate quantity to escape from the 

 dam. This was, I think, the chief cause of abnormal loss during 

 the period ; some little effect may be attributed to water consumed 

 by stock, but this could only have been visible during the periods of 

 low water, and later observations led me to the conclusion that, with 

 a larger area of water exposed, there is doubtless a larger volume 

 of water lost but a less diminution in depth as measured in the dam. 



A little consideration will show that it would have been useless to 

 keep a record in the form thus far adopted after water was dis- 

 tributed for irrigation ; close measurements were however kept by 

 myself until May, 1890, and the results given by the seasons 1887-8 

 — when we had an initial depth of 13*2 feet on March 7th, 1887, 

 and an area of nearly 2,000 morgen under water — give a clue to the 

 different rates of loss from large or small areas. The loss during 

 the twelve months March to March, 1887-8, is rather less than 

 that shown from January to December, 1887, in annexure 2 ; the 

 difference is about one-tenth of a foot. These measurements are 

 those which show the least loss, vertically measured, of any taken 

 during the 6^ years' work, while the mean area flooded has not, at 

 any time, been nearly approached. 



This result is much' greater than that obtained by Mr. Eussell at 

 Lake George, for deducting my constant (considered later on) for 

 percolation, there still .remains the remarkable difference of Lake 

 George giving up only 41 inches, against 66 lost by Van Wyks Vley, 

 and it is this discrepancy which led me to suggest a doubt as to the 

 freedom from percolation outwards or drainage inwards in the case 

 of Lake George ; there is, of course, the possibility that they may 

 neutralise each other's action. 



It being my duty to determine, each year, how much land it 

 would be safe to place under irrigation, I found it advisable to 

 determine the approximate rate of loss from percolation, and I had 

 as data — 



1. The minimum rates of total loss under different heads of 

 water. 



2. Special measurements, taken during the periods of minimum 

 loss, which were compared with mean temperatures and also with 

 differences between wet and dry bulb measurements. 



From these data curves were constructed in the hope of getting 

 at the actual loss by evaporation during these short periods. There 

 was not more discordance in the results than could be fairly 



