490 Journal of Agricultural Research voi. vi, no. 13 



In planting the seed special care was exercised in order to obtain a 

 perfect stand, and in only a few instances was there a failure to get the 

 proper number of plants for each pot. It seems in keeping with accurate 

 methods of research to plant more seeds per pot than would be required 

 for a perfect stand if they all germinated. It is safer to remove the 

 extra plants than to transplant or reseed, and the plants are more likely 

 to be uniform if it is possible to make some choice in thinning them down. 

 An exact record was kept of the number of seeds planted, and all those 

 which failed to germinate were dug out. 



For inoculating the legumes, nodules from the same crop as the plant 

 to be infected were crushed and placed in i liter of distilled water, and 

 10 c. c. of this solution were applied to the zone nearest the seed. If the 

 nodules were not available, 300 gm. of soil from a field where the respec- 

 tive legumes had been grown were well shaken with 500 c. c. of water, 

 filtered, and 10 c. c. of this solution were applied in the same manner as 

 indicated above. 



PLANT FOODS 



The first appHcation of plant food was made when the crops were 

 planted, the others at intervals of two weeks. The plant foods were 

 made up in the following manner: 



Nitrogen : Dissolved 80 gm. of ammonium nitrate, 50 gm. of potassium 

 sulphate, and 20 gm. of magnesium sulphate each in 2,500 c. c. of dis- 

 tilled water, and o.i gm. of ferric chlorid in 250 c. c. of distilled water. 

 A standard application of these plant foods was 10 c. c. of each of the 

 first three and i c. c. of the last diluted as desired. In no case was the 

 solution applied in a concentrated form. 



MOISTURE SUPPLY 



Throughout the first period of these experiments, the water content of 

 the sand was maintained at 14 per cent by weighing the jar each week. 

 This phase of the details became so burdensome that it was omitted. 

 The method was not accurate, at least during the latter period of growth, 

 because of the irregularity in plant development due to different treat- 

 ments. Some pots gave off more than 10 times the quantities transpired 

 from others. vSatisfactory results were obtained by watering the pots 

 when they needed a supply of moisture and no difficulty was experienced 

 in determining the point where the water content of the sand was below 

 normal. 



Whenever weather conditions would permit, the pot cultures were 

 placed on trucks and removed to the cage out of doors. 



TIME OF HARVESTING AND HANDLING THE CROP 



The time of harvest was governed largely by the condition of the 

 experiment. However, in most instances the same factors which control 

 the time of harvest in general farm practice held true here. The grain 



