Jnne, i92o.] SOME SOLUTION CULTURES OF WHEAT 83 



degree of apparent poorness. The physiological worth of any solution, 

 by any criterion, is thus really proportional to the reciprocal of the 

 corresponding relative score value. The solutions might be considered 

 as "first class", "second class", etc., in the ordinary sense of these 

 terms. 



Table I presents the relative score values for all the ninety solu- 

 tions without potassium, according to each of the five growth 

 criteria. The table is divided into six parts, each part representing 

 the fifteen different solutions of each type. At the extreme left of each 

 part is shown the designation of the type and the three main salts 

 used, as has been said. In the second column are shown the symbols 

 for the five criteria, as just described. The remaining' 15 columns show 

 the values for the fifteen different sets of salt proportions, these being 

 designated by the symbols already explained. At the bottom of each 

 part of the table are shown the average score values, for all five 

 criteria combined. 



Since the available data represent only a single trial, — one culture 

 of five plants for each of the ninety solutions, — the following dis- 

 cussion will be confined to a consideration of the average or generalized 

 values. These averages may be taken to represent approximately the 

 relative physiological worths of the respective solutions. The\' range 

 from 1.4, for the best incomplete solutions, to 3,4, for the poorest. Of 

 course the average for the control solution (with potassium) is 1.0. 

 The averages have been grouped into three classes : values from 1.0 ' 

 to 1.7, inclusive, are regarded as good ; those from 1.8 to 2.6 are 

 medium, and those from 2.7 to 3.4 are poor. Generalized values of the 

 good class are shown in table I by black-face type, while those of 

 the poor class are shown by Italic tj'pe. Table II presents a sum- 

 mary of the solutions of the good and of the poor class. In the 

 good class the three very best solutions (1.4) are shown by black-face 

 type (these are just alike in their scores by all criteria) and those 

 that have a score of more than 2 by any one criterion are shown by 

 Italic type. It is seen that all but the three best (11 out of 14) are 

 alike in having the generalized value of 1.6, and all of these are to 

 be considered as closely approaching the three best (1.4), those in- 

 dicated by Italics being not- as promising as the others (but see also 

 the footnotes of the table). 



