1910] +~REED—TRANSPIRATION AND GROWTH OF WHEAT 97 
The average transpiration for > unit of green weight was 56.77% 
in the controls, 34.81 for KCl, and 34-05® for K,SO,. Expressed 
in percentage of the controls, the addition of KCl caused a use of” 
Water amounting to 61.4 per cent. of the controls, and K,SO, caused 
. 4 use of water amounting to 60 per cent. of that used by the controls 
in untreated soil extract. . : 
It appears, therefore, that potassium is more efficient than sodium 
in diminishing the transpiration per unit of green growth: produced. 
Different salts of potas- 
sium did not exhibit wide 
variation in their inhibit- 
ing action upon transpira- 
tion. Sodium, however, 
is able, to some extent, to 
retard the transpiration 
per unit growth, and part 
of the retarding action of 
sodium nitrate is therefore 
attributable to the action 
of the sodium ion. 
The effect of adding 
these salts to soil extracts 
is shown by curves in figs. 
6, 7, 8, and g, where the 
method of constructing 
the curves was similar to 
that used for Pes 3, a, . 
and 5. In fig. 6, which shows the effect of K,SO,, it is evident that 
the correlative transpiration is much lower than the growth, a 
“very point in this transpiration curve is below the green weight 
curve. The middle portion of the transpiration curve lies nearer 
the green Weight curve than either of the extremities, but the general 
trend of this curve shows that constant retarding action upon trans- 
Piration in all cases whether the green weight of the plant was 
increased much or little. s 
The effect of NaNO,, as shown in fig. 7, is not ” marked = 
retarding transpiration, Here the mean of the transpiration curve 
Fic. 8.—Correlative growth and transpiration 
in soil extracts to which Na,HPO, was added. 
