252 



E. E. FREE 



that the result, as shown in the table, is a record of individual 

 comparisons of each plant with each other plant and of the judg- 

 ment formed in each case. 



To reduce this to roughly quantitative terms it is necessary 

 only to add the scores given to each plant in the course of the 

 comparison. Doing this, it is found that 1 has been scored once, 

 2 not at all, 3 three times, 4 twice, 5 six times, 6 and 7 each four 

 times and 8 seven times. 8 is, therefore, in the best condition 

 and 2, which was not scored at all, is in the worst. In figure 2 

 the results of adding the scores are plotted, the horizontal axis 

 carrying simply the numbers of the plants and the height of the 

 corresponding point above this axis being an approximate measure 



Fig. 2 



Fig. 1 



of the goodness of condition of that plant relative to the con- 

 dition of the other plants in the series. 



It is apparent that the actual vertical difference between 

 two plants next each other on the curve may not represent 

 accurately the degree of difference in condition. Thus the curve 

 of figure 1 would be the same if plant number 8 were only a 

 little better than plant number 5 or if it were a great deal better. 

 In some measure this can be corrected by a system of weighted 

 scores indicated by overscoring or underscoring the figure 

 of the score as set down in the table. The writer is accustomed 

 to use five such weights as follows. The greatest difference 

 existing between any two plants of the series is doubly under- 



