METHODS OF EXPERIMENTATION WITH COCKLEBUR 79 



placed at one end of the bench so that they do not shade the shorter 

 ones. Plants are arranged after trimming so that the leaves all face 

 in one direction and therefore do not shade each other. There is 

 admittedly an element of superstition in this, but perhaps this is 

 something to be recognized and used more often in science. We have 

 not yet carried out detailed experiments to find out how important 

 some of these procedures might be, but in the meantime we follow 

 the procedures anyway. 



In setting up the experiment, plants are numbered consecutively 

 beginning on one end of the bench, in the manner illustrated by the 

 diagram in Fig. 5-5. All of the plants with the same number receive 



©®©@ooooooooooooo©oo©® 

 ®@©@oo®oooo©oooooooo®@ 



®©®©o©o©©©©©©©©©©©©©©© 

 ®@®@©©©©©©©©©©©o©©©©©© 



©©©©©©©©©©©©©©©©©©©©©© 

 ©@®©@©©®©©©©©o©©©©©©©© 

 ®@o@@©©©©©©@©©©©©©®®®@ 



®©@®@©©©©©©0©©©0®®0®©®@ 



o©©®©®®®®®®®®®®®®®®®©®© 



TWENTY TREATMENTS, TEN PLANTS PER TREATMENT 



Figure 5-5 

 Method of numbering plants on the bench to insure that each group of 

 plants receiving the same treatment is exposed to the full range of en- 

 vironmental variability. The arrangement illustrates an experiment with 

 20 treatments and 10 plants for each treatment. 



the same experimental treatment. If in this numbering system, all of 

 the No. 1 plants end up on the front of the bench, then we change the 

 number of plants in the row or make some other adjustment so that 

 the number ones form a diagonal across the bench as shown in Fig. 

 5-5. The idea is to try to expose each experimental treatment to all 

 of the environmental variability which might be found on the green- 

 house bench. Plants in the middle of the bench always grow taller, 

 and one end of the bench may be closer to the heater, more shaded, etc. 

 This method of numbering provides a sort of "non-random 

 randomness". Since the treatments do occur along diagonal lines. 



