132 
Transactions of the Boyal Society of South Africa. 
to rise to the surface by capillary attraction, in the way already out- 
lined, and accordingly the surface soil is given such an accumulation 
of salts that cultivation is greatly hindered, if not rendered absolutely 
impracticable. 
Under these circumstances it becomes an important matter to consider 
what proportion of sodium salts a soil may contain and still be capable of 
successful cultivation. Professor Hilgard has devoted considerable time 
and attention to investigating this point in connection with the alkali 
soils of California, and he states (Bulletin No. 128): "Barley failed to 
grow where the total salts were '203 per cent., but gave a full crop where 
they were '159 per cent, (half of which was carbonate). Wheat is rather 
more sensitive ; maize fails on slightly alkaline land ; but certain sorghums 
do well on mild white alkali (i.e., not containing much carbonate)." 
In India Dr. Leather's experiments have led him to conclude that 
germination was possible where the percentage of carbonate did not 
exceed '4, nor the sulphate and chloride 1 per cent. In after-growth 
•2 per cent, of carbonate was harmful, and '4 generally fatal. Maize was 
least affected of the cereals. Wheat grew well in the presence of "137 per 
cent, of carbonate, but was destroyed by '2 per cent. Legumes are more 
affected than cereals, but even the latter were fatally affected if the soil 
around their roots contained a proportion of "2 per cent, of sodium 
carbonate. 
Methods of Analysis. 
The chemical analysis of an alkaline soil is far from being as arbitrary 
in its nature and in the method by which its results have to be interpreted 
as is the case with the investigations into the proportions of plant food in 
soils. There is therefore not always the same necessity when tabulating 
results of analyses of alkaline soils to specify particularly the methods 
adopted by the analyst. And yet it is for many reasons not only con- 
venient, but also advisable that the methods employed in the Government 
laboratories should be stated here. These methods, as now carried out, 
are usually as follows : — 
1. Preparation of Solution. — 100 grammes of true soil (i.e., the field 
sample sifted through a 3 mm. sieve) are placed in a flask marked to con- 
tain 1,000 c.c, and the flask is filled up to the mark with distilled water 
and placed in a revolving coatinuous shaking apparatus for twenty to 
twenty-four hours. At the expiration of that time the solution is either 
filtered under pressure through a dry filter placed in a porcelain perforated 
plate funnel, or by means of a Berkefeld candle filter. 
2. Detcniiination of Total Soluble Salts. — 50 c.c. of the above soil solu- 
tion are evaporated to dryness in a tared platinum dish, ignited at a tem- 
perature just below redness, so as to destroy any organic matter present, 
and weighed. 
