378 Report S.A.A. Advancement of Science. 



The data obtained from rain guages, however, are not in them- 

 selves conclusive, inasmuch as in all cases the whole of the rainfall 

 could not be collected or stored owing to the absorption by the soil 

 and evaporation by sun and wind. Hence it is necessary, in order 

 to arrive at a correct idea of the amount of water available, to 

 measure, in conjunction with the rainfall, the amount of flow-off in 

 river channels. 



This is the object of the proposed hydrographic survey, the 

 importance of and necessity for which has been impressed upon 

 successive Governments, but so far without result. 



In this Colony, where we do not derive any water from the 

 eternal snoAvs which are denied to South Africa, we are dependent 

 entirely upon rainfall and its conservation, and to our rivers we must 

 look for our supplies — yet of the capacities of our rivers we are as 

 vet absolutely ignorant. 



The hydrographic survey is necessary to esta1)lish :- - 



I. The amount of rainfall over a given area, which is called a 

 watershed, and which embraces the sources and tributaries of a great 

 river. 



::. The periods of the year during which such rainfall occurs in 

 the various sub-divisions of a watershed. 



3. The amount of flow-off at various points along the course of 

 the river, that is, the difference between the total quantity of rain- 

 water which falls and that which has evaporated, or has been absorbed 

 by the soil before the point is reached at which measurements are 

 being taken. 



As a ground work or basis upon which such observations will be 

 made, the Colony has been divided into ten distinct hydrographic 

 areas or districts, each of which comprises the whole of one or more 

 large rivers, and each such district may be sub-divided into smaller 

 districts, which may embrace one or more of the tributaries. 



The first work of the survey would consist of laying down 

 definite boundaries for each district or sub-district, and to ascertain 

 the exact superficial area of each. Rainfall statistics would be ob- 

 tained at various points in order tO' establish a fair average figure for 

 such rainfall, and at suitable points measuring weirs would be placed 

 to ascertain the amount of flow-off. Similar weirs placed at or near 

 the confluence of several tributaries would establish a check upon 

 the detailed readings, and would allow of arriving at a fair estimate 

 of the losses due tO' soakage and evaporation in the larger channels. 



Data collected in this manner would be of the greatest value 

 and utility in estimating within a comparatively short period of time 

 the potentialities of any irrigation scheme submitted to GoAemment, 

 while under present conditions much of the information required has 

 to be guessed at or based upon entirely insufficient information. Nor 

 are such data of use in the case of irrigation alone. 



In the designing of bridges, for instance, the principal item 

 of knowledge required is the maximum flood level recorded at the 

 site fixed upon. At present it is an established custom, when 



