114 CHAPTER VII 



water, that is to say as water in excess of that which can be absorbed by 

 capillarity. Hygroscopic water is not usually available for plant use, 

 and gravitational water is injurious to all except a few specialized plants. 

 Generally normal vegetative growth occurs between the limits where the 

 hygroscopic water ends and the gravitational water begins, that is to say 

 when a soil contains only hygroscopical water and a little capillary water 

 the plant will wilt, and when gravitational water is present normal growth 

 is checked. Experiment has shown that usually plants will make their 

 maximum growth when the maximum quantity of water is present that can 

 be absorbed by capillary attraction. The actual percentage of water in 

 a soil corresponding to this condition varies within wide limits ; thus in 

 sandy soils the hygroscopic water is about 2 per cent., rising to 10 per cent, 

 in clays, and to 40 per cent, in peats ; the actual water content for the best 

 results will be least in sandy and most in peaty soils. This feature of irriga- 

 tion has been studied to some extent by Eckart^^ and more recently by 

 Burgess.^* The latter calls attention to the very hygroscopic nature of 

 Hawaiian soils due to the presence in large amounts of colloidal silica, ferric 

 oxide, alumina and humus, and he estimates that soils such as these are in 

 the optimum condition when they contain about 45 per cent, of water. 

 Eckart, experimenting on the soils of the Experiment Station in Honolulu, 

 found that the best results were obtained with an irrigation of three inches 

 per week, the soil then containing on an average 31-38 per cent, of water. 

 As this soil could absorb 40-74 per cent, of water, the optimum percentage 

 would occur when it was saturated to 77 per cent, of its capacity, a figure 

 higher than is found with most crops. 



Quality of Irrigation Water. — MaxwelP^ arbitrarily fixed the " danger 

 point " of irrigation water at 100 grains of salt per imperial gallon ; Hilgard^^ 

 states that 40 grains is the usual limit. Eckart^ ^ found cane in lysimeters 

 grew unchecked when the soil water contained 195 grains chlorine, as 

 sodium chloride, per U.S. gallon, and obtained in lysimeters a normal growth 

 when irrigation water containing 200 grains of salt per gallon was used in ex- 

 cess, at the same time permitting good drainage from the porous soil em- 

 ployed in the tests. He also found that gypsum and coral sand mitigated 

 the harmful effect of saline irrigation waters. ^^ 



The nature of the salt in the water has a profound effect ; sulphates or 

 carbonates of lime and magnesia are not harmful ; it is in the chlorides of 

 the alkalis that danger lies. The danger of such water lies in their abuse 

 rather than in their use ; if the soils to which they are applied are ill-drained 

 so that the salt can accumulate, the quantity soon becomes toxic ; combined 

 with natural rainfall, applications of a purer supply or heavy applications 

 of the saline water, together with good drainage so as to wash out the 

 accumulated salt, permit their safe use. 



Conservation of Soil Water. — After the water has arrived in the soil 

 a great part is always lost by evaporation, and this is capable of control 

 within certain limits. A protective layer of soil in fine tilth prevents the 

 upward movement of the water by capillary attraction to the surface, and 

 is highly efficient in retaining water in the soil. Not less important is the 

 nature of soil ; soils containing much humus are especially water-retentive, 

 and this is capable of control by burying the trash of the cane and by 

 ploughing in green manure ; to a certain extent the benefits of these 



