138 Transactions of the Boyal Society of South Africa. 
It would appear that the most noxious of the sodium salts, the car- 
bonate, is practically absent from the Houw Water soils, the two com- 
pounds of sodium which form the bulk of the alkaline soil constituents 
being the sulphate and the chloride. Professor Hilgard estimates that, 
for barley, sandy soils should not contain within the first 12 inches over 
•10 per cent, of sodium carbonate, -25 per cent, of sodium chloride, or 
•45 to -50 per cent, of sodium sulphate ; in clayey soils even smaller per- 
centages may prove injurious. It has been calculated that the average 
proportion of total alkaline salts in the 4 feet nearest the surface should 
not exceed -2 per cent, if barley is to be grown. According to Professor 
Hilgard's experiments, as ah^eady stated, barley thrived on a soil which 
contained on an average -159 per cent, of alkaline salts in the first 4 feet ; 
on the other hand, the cereal refused to grow where the average amount 
in the first 4 feet reached -203 per cent. Now, of the Houw Water soils it 
seems likely that A would satisfy this condition, and possibly B may also ; 
the average percentage in C is -186, but both D and E are considerably on 
the worse side of the danger limit." 
Division of Steynsbueg. 
A similar investigation to that outlined above was made at Thebus, in 
the Steynsburg Division, a distance of about 160 miles E S.E. of Houw 
Water. In this case, too, there was a proposal afoot to construct a dam 
whereby the farm Zout Kuil and part of Van Yuurens Kraal with adjacent 
lands could be irrigated. As Parliament had voted £150,000 for the 
purpose, much depended on the suitability of the area. The last-named 
farm had been under cultivation for close on to half a century, and the 
occupant prided himself on the returns which he has obtained by means of 
irrigation. At a point within this cultivated area a set of samples was 
taken in order to ascertain, if possible, the effect of continuous irrigation 
on the soil, by comparing the analytical results yielded by the irrigated 
soil at this spot, which will be called K, with those of samples from the 
adjacent tracts which had never yet been irrigated. This point, K, was 
on the opposite side of the railway to that where the dam construction 
works w^ere situated, and is rather more than 3 miles below the dam, and 
1-| mile from Thebus railway station ; it constituted, in fact, the head of 
the area which it was intended to irrigate. Four samples, L, M, N, and P, 
were taken at other points in this area, on the farm Zout Kuil, extending 
altogether over a space about 2 miles in extent. 
The alkaline salts in these five samples were determined in similar 
manner to that employed with regard to those from Houw Water, only 
much more completely. 
In this way the following percentage results were obtained : — 
* ^ It is of course obvious that efficient drainage may prove a complete safeguard. 
