10 Transactions of the South African Philosophical Society. 



to be assured that no underground drainage enters the lake in dry 

 seasons, for I understand the admitted drainage in wet seasons to 

 mean underground drainage ; without reasonable certainty on these 

 points a valuable factor in making comparisons is missing. 



I will now sketch the steps taken to distinguish evaporation from 

 percolation at Van Wyks Vley. The reservoir was completed about 

 December, 1883, and very soon after took water ; the resident 

 engineer, Mr. J. E. McNellan, under Mr. J. G. Gamble, then 

 Hydraulic Engineer to the Colony, began to register the depths of 

 water on the 23rd of January, 1884, and kept the record until the 

 28th of October of that year, when I made a survey of the area 

 (at level of sill 33 acres), then flooded, with the water 3 inches 

 above the sill upon which the discharge pipe rests. After this 

 date I often supplemented Mr. McNellan's observations with others 

 taken by myself, using the same measuring-rod and position, and 

 we continued this record until the 17th of June, 1885, after which date 

 water was used for irrigation purposes. The point from which 

 measurements have throughout been taken is that reached by 

 passing a rod, vertically held, close up to the centre of the discharge 

 valve and down to the stone sill upon which it rests. The rod was 

 of mahogany, painted and divided to hundredths of a foot. It will 

 at once be seen that measurements taken in this way cannot be too 

 critically compared with the tank observations taken in New South 

 Wales, and the second decimal places I use in the tables which I 

 will annex to this paper are only useful in so far as they assist in 

 checking and balancing total results. 



The data given in Table I. were gathered in a period of unusual 

 drought, the rainfall for the whole period of nearly eighteen months, 

 including two rainy seasons, being only + |*|g} 8*08 inches, and the 

 results reached may safely be taken to represent the maximum 

 vertical loss by evaporation, percolation, and waste in spray taken 

 together that is ever likely to be experienced at Van Wyks Vley. 

 These data are not, however, to be relied upon as indicating the 

 present rate of loss, for several reasons. When I made the little 

 survey before alluded to I determined the depth of the original 

 water table, underlying the dam, with reference to the point from 

 which the depth of water within the dam is measured, and I found 

 it to be about 7 feet below that point. In constructing the dam 

 the contractor carried his excavations in search of material (within 

 the dam) nearly down to the level of this water table, and, during 

 the whole of the first year, soakage through the thin crust left was 

 sufficient to cause little whirlpools to show themselves on the 

 surface of the water, indicating a very rapid escape from the dam 



