io6 Journal of Agricultural Research voi. xviii, no. 2 



usually impracticable, so it is necessary to approximate the requisite 

 condition by changing the solution at more or less frequent intervals. 

 In order to determine exactly what has been the average condition of 

 the solution between changes, analyses must be made as in the pre- 

 viously described experiments. Obviously the extent of change in the 

 solution will depend upon the rate of absorption by the plant and also 

 upon the concentration and total volume of solution per plant. Con- 

 sequently, actual determinations of the quantity of each ion absorbed 

 must be the basis for the selection of suitable culture vessels, number of 

 plants, and times of changing solution. If too small a total supply per 

 plant of any element is present, all or nearly all the ion in question may 

 be removed from the solution in the interval between changes of solu- 

 tion. Since all of the ions are not absorbed in equal percentages, not 

 only the total concentration of the solution but the ratio between ions 

 will be changed. The average composition of the solution will depend 

 then upon the particular set of empirical conditions chosen. How very 

 important these considerations are is indicated by all the absorption 

 studies of this investigation. It will be recalled, for example, that when 

 two barley plants seven weeks old were placed in 1 liter of a solution of 

 about 200 parts per million total concentration, in less than 72 hours 

 every trace of NO3 and over 90 per cent of the K and PO4 were removed 

 from the solution. It is quite clear from such an experiment that abso- 

 lute quantities rather than concentration may have been the limiting 

 factor. In fact a later experiment indicated that such was the case. 



Brenchley (5) grew barley and wheat plants for seven weeks in solu- 

 tions of 3,000, 600, 300, and 150 parts per million. The solutions were 

 changed every four days. She concluded that the lower concentrations 

 are sub-optimum and criticised Stiles' (43) results. The latter grew 

 single plants in 1,200-cc. bottles for six weeks, changing the solution 

 every three or five days and using concentrations of 1,750 parts per 

 million and one-fifth, one-tenth, and one-twentieth of that concentra- 

 tion. He did not find a significant difference in yield, although there 

 was some falling off in the lowest concentration. 



Tottingham (45), as a result of various experiments by the water- 

 culture method, using 250- or 400-cc. bottles with six plants to a bottle 

 for a growth period of 23 days, decided on a concentration of 2.5 atmos- 

 pheres as optimum, though he recognized that a less concentrated solu- 

 tion might give an equal yield. 



Shive (39), with a technic similar to that employed by Tottingham 

 except for the use of a 3-salt nutrient solution, came to the conclusion 

 that o. I atmosphere was sub-optimum concentration and assumed the 

 optimum concentration to be 1.75 atmospheres. This concentration has 

 since been adopted by other investigators as optimum. In Shive's 

 (40, 41) sand-culture experiments about 250 cc. of nutrient solution 



