A RELATIVE SCORE METHOD 251 



without, however, any implication that it involves principles 

 essentially new. 1 



In essence the method consists in the individual recording, 

 and later summation of each-to-each comparisons between each 

 two individuals of the series being studied. It will be clearer by an 

 example. The data are of an actual case of comparison between 

 eight plants as to their general health and condition and are 

 taken from an investigation now under way in this laboratory. 

 The first step is to prepare a ruled sheet as shown in figure 1, 

 the numbers of the last seven plants being set down in the upper 

 margin, one to a column, and the numbers of the first seven in 

 the left hand margin, one to a line. Plant number 1 is then 

 compared with plant number 2 as to their relative condition — ■ 

 all other plants in the series being neglected. It is judged that 

 plant number 1 is better. Accordingly in the space in the 

 table corresponding to plant number 1 on the left and to plant 

 number 2 on the top there is set down a "1" this constituting 

 a " score" for plant number 1 and indicating that in that partic- 

 ular comparison plant number 1 was judged better. Proceed- 

 ing, the comparison of plant number 1 with plant number 3 

 leads to a judgment that number 3 is better and a "3" is set down 

 in the square to the right of the "l' ; first marked. Similar 

 comparisons of plant number 1 with plants numbers 4, 5, 6, 7 and 8 

 indicate that in each case these are better than number 1 and 

 the appropriate score is set down on the proper square. AYhen 

 plant number 1 has been compared with all the other plants in 

 the series, plant number 2 is compared with each other plant 

 (except number 1, this comparison having been made already), 

 and the appropriate scores set down in the second line of the 

 table shown. Plants numbers 3, 4, 5, 6 and 7 follow in turn 

 with the results shown in the table. When plant number 6 is 

 compared with plant number 7 no difference can be detected 

 and an "a," is therefore set down, meaning that the plants are 

 alike and counting as no score for either plant. It will be noted 



1 A brief description has already been published by the writer, Jour. Physical 

 Chem.13: 131 (1909). 



