1918] Waynick: Antagonism and Cell Permeability 143 



obtained by other investigators in comparable periods of time. A 

 further discussion of this point will be taken up below. 



At the expiration of the six weeks growing period the plants 

 were removed from the corks, the roots rinsed thoroughly with dis- 

 tilled water, placed between layers of paper toweling, dried in the 

 oven at 100°-105?C, roots and tops separated, weighed, and placed 

 in envelopes ready for analysis. For analysis the roots from dupli- 

 cate cultures were combined unless the dry weight was sufficient to 

 allow of separate analysis. 



Total ash was determined after direct ignition of the dry material 

 in a muffle at a low red heat until no trace of carbon remained. The 

 ash was then taken up in dilute hydrochloric acid and evaporated 

 to dryness to remove possible contamination with silica. Iron was 



precipitated as the hydroxide with ammonia and titrated with 



100 

 potassium permanganate after reduction with zinc and sulphuric acid. 



This determination was made because of the relation Gile has shown 



to exist, between calcium and iron absorption by plants. Calcium 



was precipitated as oxalate and titrated with — — potassium perman- 

 ganate. The double precipitation of the oxalate assured freedom 

 from magnesium contamination. Magnesium was precipitated by 

 ammonium phosphate and weighed as the pyrophosphate. Potassium, 

 where determined, was precipitated and weighed as the chloroplati- 

 nate. Copper was determined colorometrically by using the Eerro- 

 cyanide method. The amount of material available precluded the 

 possibility of a more complete analysis than was made if any degree 

 of accuracy was desired. For example, in Series vn, the weight of 

 the ash varied from 12 to 233 milligrams in the case of the roots and 

 from 32 to 183 milligrams in the case of the tops. While these varia- 

 tions are not extreme, they are fairly representative. The values of 

 these elements actually determined cannot be taken as absolute in 

 every case because of the limited amounts of material available, but 

 the significant differences are so great as to make a small variation 

 in this regard of minor importance. 



The strength of all solutions is uniformly expressed in terms of 

 molecular concentrations since this mode of expression has been 

 quite generally used in experimental work reported by different 

 investigators. 



Under experimental results twenty-six series are reported. A 

 series, as used in the present work, may be defined as a number of 



