June, 1920.] SOME SOLUTION CULTURES OF WHEAT 37 



monly set up for elementary students of plant physiology, to show the 

 effect of the omission of one or another of the essential chemical 

 elements, must be regarded as quite worthless. For such, demon- 

 strations it is logically essential that the incomplete solution used be 

 the very best one possible without the omitted element. The further 

 development of physiology and fertilizer practice in agriculture requires 

 that the sort of work here preliminarily presented should be carried 

 much further. A next step along the path of this study might well be 

 to carry out experiments similar to those here described, including only 

 the solutions charactetized as belonging in the good class, and to 

 continue the cultures until the best ones show marked injury or cessa- 

 tion of growth. It would be desirable, furthermore, to test these sets 

 of salt proportions with total volume-molecular concentrations different 

 from the one used in our series; it seems, especially, that more dilute 

 solutions might have greater physiological worths than any tested by 

 us. It would also be desirable to renew the solutions in the cultures 

 at much more frequent intervals, or, better still, to arrange the cultures 

 so that the solutions would be constantly renewed. 



A rather thorough study of physiological balance in this series of 

 solutions was originally contemplated, but the fact that the work had 

 to be discontinued at the end of the single preliminary series makes it 

 undesirable to attempt a detailed discussion of the relations between 

 the salts used and salt proportions, on the one hand, and the growth 

 values, on the other. The numerical data obtained are very consistent 

 among themselves in many ways, however, and some of the most 

 striking observations in this connection may be added. 



All of the plant data presented in table 1 were placed upon 

 triangular diagrams of the form used in planning the solutions, and 

 these were inspected for relationships between growth and salts and 

 vsalt proportions. Only the six diagrams for generalized score values 

 (averages, table I) will be considered here. These six diagrams are 

 shown in figure 1. In studying these it is to be remembered that any 

 set of salt proportions represents exactly the same proportions of the 

 atomic groups (N0 3 ) 2 , (H 2 P0J 2 and S0 4 on all six diagrams. A given 

 set of salt proportions differs from one solution type to another only 

 (1) in the proportions of Ca and Mg, and (2) in the way these two 

 potential cations were combined with the three potential anions in the 

 salts used. On the diagram of figure 1 the three very best solutions 

 (generalized score, 1.4) are each shown by a heav}^ circle. The remain- 

 ing solutions of the good class as above characterized (1.6) are shown 

 by lighter circles, those designated by Italic type in table II having a 



