July, 1911.] 



65 



Miscellaneous, 



An interesting case of this nature is 

 found in Mannar island, where the 

 writer had cn two occasions to make 

 investigations for water supply. At the 

 eastern (Mannar town) end of the island 

 and on the south-western shore the 

 winds of the south-west monsoon have 

 thrown up sand dunes, and these 

 extend inwards for, roughly, one-third 

 ot the width of the island and lie over 

 the subsoil coral formation. The nat- 

 ural escape for the rain water by percol- 

 ation is situated all along the shore 

 margin, and the limiting level is that 

 of mean sea. The surplus water after 

 rains flows out to sea until the water 

 plane takes its limiting slope, and fresh 

 water wells can be sunk in the sand 

 along the shore almost up to the wash 

 of the waves. In addition, the fresh 

 water surface in these wells actually 

 rises and falls with the tides, as the 

 flow is temporarily arrested (or decreas- 

 ed) at the time of high water. 



The inland margin of this sand dune 

 tract borders on a silty plain, where the 

 percolation is partially arrested, so the 

 best position (i.e., the position where 

 the greatest depth of fresh water can 

 be obtained) for wells is on the inland 

 border of the dune tract. 



Similar conditions prevail at the other 

 end of the island (Talaimannar), but 

 there the whole width of the island is 

 made up of sandy soil, and so, in plot- 

 ting a contour map of the subsoil water 

 levels, they will be found to represent 

 curved lines roughly parallel to the 

 coast line, and culminating in a point 

 which is equidistant from the shore on 

 the north, south, and west sides. This 

 point represents the position of best 

 supply from a well. 



Before leaving this subject it may be 

 mentioned that the loss by evaporation 

 from the surface of sandy soil is only 

 about one-fifth of that from the surface 

 of water, and the actual relative losses 

 by evaporation from water, earth, sand 

 and other surfaces may be taken as 

 approximately in the following ratios :— 



Grass sod surface ... 192 



Cereals ... ... 1-73 



Forest ... ... 151 



Water ... ,.. i-00 



Bare soil ... ... 65 



Sand ... ... 0-20 



Soil covered with forest leaves 0'07 



It will be seen from these figures that 

 the losses due to evaporation from sandv 

 surfaces are only about one-third of 

 those from bare earth surfaces, and 

 about one-tenth of those from grass 

 lands. As " evaporation " and the 

 losses classed under this heading may 

 9 



account for 60 to 70 per cent, of the 

 rainfall, it is important to bear this 

 point in mind in the selection of well 

 sites, having regard to the conservation 

 of the available subsoil water supply. 



Incidentally also this matter is of con- 

 siderable interest to agriculturists in 

 the consideration of the question as to 

 what condition their ground surfaces 

 should be kept in. 



Well tilled, loose earth surfaces re- 

 semble the sand formation, and greatly 

 decrease losses by evaporation. In the 

 drier districts the tea planter who keeps 

 his ground surface free of all grass and 

 weeds and the soil loose would appear 

 to be acting wisely in so doing, as he is 

 helping to conserve the water supply 

 for the roots of his plants. But this is 

 perhaps a digression. 



10. In the above paragraphs the ex- 

 treme conditions have been dealt with, 

 but by far the most difficult problem 

 remains in the case of selection of well 

 sites in districts where the ground sur- 

 face is all composed of ordinary soils, is 

 only very gently sloping, and gives no 

 special indication of the subsoil form- 

 ation. There may be no rock and no 

 sandy soils for miles round. Speaking 

 generally, such localities are not pro- 

 mising, and in many places the only 

 course left is to take the risk and make 

 a trial by either borings or pits. 



When the writer was attending lec- 

 tures in the University School of Engi- 

 neering, an old Professor once made the 

 remark ( probably repeated to each 

 successive class of students ! ) " always 

 look out carefully for other men's fail- 

 ures as such failures are only unintended 

 experiments." This is pretty sound 

 advice, and it may be repeated here. 

 In these difficult districts, where the 

 surface of the ground gives no help 

 in selecting a well site, the best course 

 is to hunt round and find, if possible, 

 some other person's success or failure. 

 In the former case the water plane 

 is in view, and the probability is that 

 it extends some distance round, though 

 this is not always true. If two or more 

 existing wells show the water plane 

 at moderate depths, then it is fairly safe 

 to try a site in some position between 

 them, and in lower ground if possible. 

 If the water planes in two or more wells 

 stand about the same level, the chances 

 are all the better. Failing any existing 

 wells, some indications may be given by 

 examining the river or stream beds or 

 water-courses in connection with a good 

 topographical map (if such exists) of the 

 district and the records of the nearest 

 rainfall gaugings. If these water-cour- 

 ses are unduly numerous and their crosa 



