[Vor. 7 
308 ANNALS OF THE MISSOURI BOTANICAL GARDEN 
the bases required are all soluble to a high degree. Tf, there- 
fore, nitrogen is furnished as NO,, the salt furnishing this ion 
would necessarily be added periodically, and to this extent the 
concentration would change from day to day. Relatively in- 
soluble salts of ammonium are obtainable, however, such as 
MgNH,.PO, and this salt has been employed in some of the 
experiments. 
In the various experiments which have thus far been under- 
taken the sources of Ca are as follows: CaSO,+2H.0, CaCO,, 
Ca,(PO.)», and CaHPO,4-2H.0; of Mg: Mg80,+7H:0, MgCO,, 
Mg,(PO.):+8H.0, and MgNH,;PO,4-6H.0; of K: KNO, and 
K,PO,; of Fe: FePO,+4H.0, FeC.0,+2H.0, ferric citrate, 
and “soluble ferrie phosphate." Certain other salts which 
might have been employed to advantage were not available at 
the time. 
Among the more soluble of the salts included in this category 
are CaSO,+2H.0 having a solubility of 0.241 and 0.222 in 100 
parts of water at 0° and at 100? C. respectively, and among the 
more insoluble salts are FePO,+4H.0 and CaCO.. One of the 
chief reasons for burdening the experiments with such a variety 
of substances may be found in the fact of their diverse solu- 
bilities; and inasmuch as the antagonistie relations of the ions 
in respect to the plant require consideration and are involved 
with osmotie and nutritive relations, such a variety of combina- 
tions was necessary in order to feel assured that some of the 
results obtained might be among the most favorable that such 
types of combinations could yield. 
It is obvious that at no instant is the exact concentration of 
any salt known in such combinations, except approximately, 
in cases where soluble salts were added. However, the total 
concentration is readily determinable, likewise the electrolytie 
conductivity of the solution. One may also estimate the partial 
concentrations. When soluble salts are employed in nutrient 
solutions, the proportion of the ions is disturbed from the mo- 
ment that the roots are introduced, and there is a progressive 
decrease in concentration until the solution is renewed.  Like- 
wise the differences in the proportion of the ions, determined, 
of course, by the differential absorption rates, are not continually 
reestablished by any form of “reserve.” It would, of course, 
