THE BROADBALK WHEAT SOILS. 
119 
therefore, while only from 4 per cent (grain) to 9 per cent (straw) 
larger, have contained nearly 4<» per cent more potash. 
One is somewhat tempted to draw the inference that, in the produce 
of the sodium sulphate plat (12), soda has partially replaced potash. 
In an elaborate study of the vegetative conditions and of the ash 
constituents of the mixed herbage of grass land, embodied in a recent 
paper by the late Sir John Lawes and Sir J. Henry Gilbert, 1 we are, 
however, warned against the temptation to assume that, even in adverse 
circumstances, soda can functionally take the place of potash. It is 
pointed out that, in detect of sufficient potash, more of soda or of time, 
Or of both, will be taken up — probably as carriers of nitric acid — and 
retained by the plant, bill that the herbage will !><■ more leafy and 
immature than where abundant potash is available. The free devel- 
opment of carbohydrates functionally demands the presence of an 
adequate quantity of potash; and, if the plant is stinted in potash, 
neither phosphates nor nitrates, even with the assistance of abundant 
soda and lime, can enable it to develop to its utmost. In the Last six 
years of the fifty years now under review the annual produce of plat 
13, with potassium sulphate, has exceeded that of plat I- (with sodium 
sulphate) !>y '![ bushels of grain and hundredweight of straw. 
Both ripe grain ami stra^ consist mainly of carbohydrates, and the 
excess virtually Indicates the grea ter carbohydrate elaborative power 
conferred on the wheal plant by supplying it freely with potassium 
salts, and the inability of sodium salts to functionally replace them. 
But the output of potash from plat 1- in quit*' recent years is so much 
greater than from plat 11 without either potash or soda thai there 
can be little doubt of the practical act ion of t he sodium salts as a 
solvent of the soil potash. 
Plat 14 is similar in every way in its history to plai L2, except that 
it lias been annually dressed with 280 pounds of magnesium sulphate 
per acre instead of with :>»'.»;.l pounds of sodium sulphate, [ts yield, 
though less than that of plat has been, on the average of fifty 
years, slightly better than that of plat 1-, over half a bushel of grain 
and nearly 1 hundredweight of si raw per acre being the annual advan- 
tage. In later years the inferiority to plat 13 has become more marked; 
but the superiority to plat \-2 is still distinct, especially in the yield 
of straw. While receiving slightly less potash in its early dressings 
than plat 12, plat 14 has given in its crops nearly 100 pounds more 
potash in fifty years. There should, by calculation, be now a deficit 
of 92 pounds of potash per acre as compared with plat 11. 
The surface soil of plat 14 yields much less cit i ic-acid-soluble potash 
than was the ease in 1881, but the second and third depths still main- 
tain their superiority, leaving, as compared with plat 11, a balance 
to the good of 7G pounds per acre of easily soluble potash in the 27 
'Philosophical Transactions of the Royal Society (England], series B, 1901, 
vol. 198, pp. i:; ( J-210. 
