18 Journal of the Department op Agriculture, 



chang'es of water, air, and temperature. In districts of plenty of 

 rain the salts thus formed are dissolved as the rain sinks through the 

 soil and are eventually carried to the g-round-water and out into the 

 natural drainage channels of the region, i.e. the rivers. The amount 

 of salt formed is, relatively speaking, A^ery small even in the course of 

 a whole year ; but such as it is, it gets carried out into the rivers and 

 finally into the sea. The amount of water in the sea does not increase 

 with time ; because the sea loses as much water by evaporation as it 

 gains in rain and from the rivers which flow into it. The water 

 evaporates, but the salts are not able to do this and are therefore left 

 behind in the sea. Thus we may presume that the sea was originally 

 fresh water, but that it has gradually become salt at the expense of 

 tlieland. 



Xow let us consider the case of a rainfall Avhich is not sufficient 

 to cause drainage from the soil. Evidently the salts which are formed 

 in the soil must remain there, so that soil which in the beginning was 

 not brak has become brak through there not having been sufficient rain 

 to cleanse it. Now among the salts which are gradually formed in 

 soils are some which serve plants for food. Therefore in arid lands 

 not only do tlie soils contain brak, they also contain considerable 

 accumulations of available plant food. That is why our Karroo soils 

 are so much more fertile than, say, our coastal soils. 



Lands of the arid interior are therefore more or less brak, and, 

 furthermore, comparatively rich in plant food, because the rainfall 

 has not been sufhcient to cause drainage from them. Generally 

 speaking the soil contains more or less brak wherever it is necessary 

 to irrigate to grow crops. Brak waters, however, may be found in 

 humid as well as in arid regions. 



Distribution of Beak in the Soil. 



The brak of any given soil may be concentrated at one or more 

 levels, or it may be somewhat evenly distributed throughout the depth 

 of the soil. Whatever the distribution be it has been determined by 

 several factors, among which are (a) the nature of the soil, (b) its 

 vegetal covering, and (c) the rainfall. Irrigation tends to alter the 

 distribution of brak in the soil. Its action in this respect may be 

 anything from good to bad. 



I now propose to examine from the brak jioint of view some of 

 the more frequent soil conditions. 



(a) 71if' Soil is Freely Pervieahle anil the Underdraiiiage is Good. 



If sufficient water be used to cause a general movement of water 

 throvigh the soil and out the latter will be gradually freed of any brak 

 it originally possessed. When these are the conditions the irrigator 

 seldom experiences trouble. If, however, he has not sufficient water 

 to cause adequate drainage he mav experience the dreaded "rise of 

 brak." 



(b) The Soil is Pervieahle hut the Drainage is Bad. 



It is evident that unless some artificial system of drainage is 

 supplied the application of water cannot remove any brak from the 

 soil. On tlie otiier hand, the application of water in the absence of 



