KNOP'S NUTRIENT SOLUTION 
453 
relations have been disclosed by the results obtained, which seem to 
justify their publication as matters of general interest. 
Methods of Investigation 
Young seedlings of Chevalier barley and Little Gem peas were 
grown, according to the method previously described by Tottingham 
(5), in pint jars containg Knop's solution^ of about 0.3 percent total 
concentration in which had been incorporated a definite amount of a 
watery suspension of the solid to be studied. Six plants were grown in 
each jar, and each modification of the solution was employed in dupli- 
cate. The substances added to the usual nutrient solution were, 
respectively: silicic acid, ferric hydrate, and carbon black. 
The silicic acid was prepared by adding slowly, with constant 
stirring, the required amount of dilute HCl to a cold, dilute solution 
of sodium silicate. Preliminary tests had shown this to give the most 
satisfactory material. The gelatinous precipitate was washed with 
water, by decantation, until the clear washing fluid gave no test for 
chlorides. Ferric hydrate was prepared by dissolving ferric oxide in 
hot, dilute nitric acid. The ferric nitrate solution thus formed was 
diluted, precipitated cold with dilute ammonia water until neutral to 
litmus, and the precipitated hydrate was washed by decantation until 
free from nitrate. "G-Elf" brand of carbon black was washed with 
dilute hydrochloric acid and then repeatedly washed with hot water 
until the wash water no longer gave visible evidence of impurities 
resembling oil or tar. In each case the solid was allowed to settle 
over night in water; then the clear, upper layer of water was siphoned 
away and the actual concentration of the thoroughly agitated sus- 
pension of each substance was determined by evaporating aliquot 
portions. In view of the uncertainty regarding the composition of the 
precipitated ferric hydrate, as noted by Mendeleeff (6), the iron was 
determined as Fe203 by ignition, and from this the equivalent amount 
of Fe(0H)3 was calculated. The evaporated residue from the silicic 
acid suspension was ignited and weighed as Si02. The nutrient solu- 
tions for each jar were made up as needed from stock solutions of the 
1 The composition of the solution is shown by the number of cc. of M/2 solution 
of each salt used per liter of solution: Ca(N03)2, 20.9; MgS04, 7.1; K2HPO4, 2.5; 
KH2PO4, 2.5; KNO3, 8.5. Equal parts of KH2PO4 and K2HPO4 were used for the 
purpose of approximating neutrality more closely than obtains where but one of 
these salts is used. Four drops of colloidal ferric phosphate per jar were added with 
each fresh portion of solution. 
