9 



from flowing over the surface of the soil while the remain- 

 der of the ban^.et i';as being ooatod after filling. It also 

 prevented the water which was added from time to time 

 during the experiment from flowing off at the sides of the 

 baslcet. The baskets were filled within three-fourths of 

 an inch of the top with soil which had been thoroly mixed, 

 pulverized, and screened to remove the lumps and stones. 

 They were then dipped in paraffin until coated sufficient- 

 ly to prevent loss of water. The paraffin also served to 

 cement the soil particles on the outer portion of the soil 

 coiumn to the sides of the basKet, thus preventing the 

 formation of an air space between the soil and basket. 



The seeds were soaked in water before planting until 

 germination had commenced. This method of treatment hast- 

 ened germination and also enabled a selection to be made 

 of seeds which were of nearly uniform vigor, thus elimina- 

 ting to some extent extremes of individual variation. A 

 few more seeds were planted than the number of plants re- 

 quired. This permitted another selection to be made after 

 the plants had begun to grow. \men the plants were 2 to 

 2 1/2 inches in height, they were thinned to the required 



