76 Journal of Agricultural Research voi. xviii, No. a 



(Ca(N03)2), potassium nitrate (KNO3) , magnesium sulphate (MgS04), and 

 potassium phosphate (KH2PO4 and K2HPO4) were made and diluted 

 to the desired degree. The composition of the nutrient solutions was 

 then determined by analysis. 



In most investigations on plant nutrition, the question of the iron 

 content of the nutrient solution has not received sufficient consideration. 

 Gile and Carrero (12) have very thoroughly examined this matter and 

 have reached the conclusion that in the absence of special precautions, 

 it is often possible that the plant may be inhibited in its growth by an 

 insufficient supply of available iron. Our experience is entirely in 

 accord with this view. The presence of sufficient dissolved iron in the 

 culture solution will depend upon the form and quantity of the iron salt 

 used, upon the concentration and reaction of the solution, and upon the 

 time of standing. By direct qualitative tests with potassium sulpho- 

 cyanid (KCNS) it is easy to show that in certain cases no dissolved iron 

 is found after a comparatively short time, even though considerable 

 quantities of iron chlorid (FeClg), iron sulphate (FeS04), etc. had been 

 added to the solution. The solubility is of course greater in solutions of 

 higher hydrogen-ion concentration; but even in acid solutions, when the 

 total concentration of phosphate (PO4) is high, all the iron may be precipi- 

 tated. Iron citrate and tartrate seem to be the most effective forms of 

 iron to use. It is also desirable to add the iron solution or suspension to 

 each culture bottle at the time the solution is changed. In any case 

 there must be assurance that iron is not a limiting factor, as it might 

 easily become in solutions of higher concentrations even when the more 

 dilute solutions were plentifully supplied. The experience of Gile and 

 Carrero (12) and actual tests with KCNS on the filtered solution would 

 seem to make possible the necessary control. 



METHODS UvSED IN SAND-CULTURE EXPERIMENTS 



The method of sand culture has been extensively employed because 

 such a medium obviously affords a more natural habitat for the plant 

 roots than solution cultures can furnish. The selection of the sand is not 

 a matter of indifference, although in many experiments beach sand has 

 been used. This could scarcely be regarded as an entirely inert substance 

 free of soluble material. For the present experiments "Ottawa" sand 

 was selected. This sand is exceptionally pure. Analysis showed that 

 it contained 99.8 per cent silica (SiOj). In the latter series of experi- 

 ments the sand received additional purification by treatment with 

 hydrochloric acid (HCl) and subsequent washing with water. Unless 

 the sand is treated ^vith HCl it is possible that a trace of alkaline-reacting 

 substance may alter the reaction of certain nutrient solutions. After 

 the treatment described above, however, it has not been found that any 

 significant change in reaction takes place as a result of contact with the 



