Methods of Working 13 



optima in this respect. For several years a solution of medium strength 

 was used, containing the following : 



Potassium nitrate I'O gram 



Magnesium sulphate ... ... '5 



Sodium chloride ... ... ... '5 



Calcium sulphate '5 



Potassium di-hydrogen phosphate '5 



Ferric chloride '04 



Distilled water to make up 1 litre. 



This is an excellent solution for barley plants, giving good and 

 healthy growth. While peas grew very well in it, they showed some 

 slight signs of over-nutrition. A weaker solution is being tested which 

 gives very good results. Peas grow very strongly in it and it also 

 seems to be sufficiently concentrated to allow barley to carry on its 

 growth long enough for the purposes of experiment. The solution is 

 as follows : 



Sodium nitrate '5 gram 



Potassium nitrate ... ... *2 



Potassium di-hydrogen phosphate. '1 



Calcium sulphate '1 



Magnesium sulphate ... ... *1 



Sodium chloride '1 



Ferric chloride '04 



Distilled water to make up 1 litre. 



The latter solution was made up so that the quantity of phosphoric 

 acid and potash approximated more or less to the amount of those sub- 

 stances found by analysis in an extract made from a good soil. 



The experiments are usually carried on for periods varying from 

 4 10 weeks, six weeks being the average time. Careful notes are 

 made during growth and eventually the plants are removed from the 

 solutions, the roots are washed in clean water to remove adherent food 

 salts, and then the plants are dried and weighed either separately or in 

 sets. In order to reduce the error due to the individuality of the plants, 

 five, ten or even twenty similar sets are grown in each experimental series, 

 the mean dry weight being taken finally. Also the same experiment is 

 repeated several times before any definite conclusions are drawn. 



Another method of water cultures is used by some investigators, 

 in which the experiments only last for a few hours or days, usually 

 24 48 hours. While such experiments may not be without value for 

 determining the broader outlines of toxic poisoning, they fail to show 

 the finer details. The effect of certain strengths of poison is not 



