BACTERIA AND ENZYMES IN AGRICULTURE 263 



multiply with great rapidity, since they are now free from 

 the attacks and the competition of their enemies, the other 

 large organisms. The dead organisms, in fact, were shown 

 to afford food for the bacteria. 



It was further found that plant growth increased in 

 partially sterilised soil, although nitrification was inhibited; 

 under these conditions it appears that the plants can obtain 

 their nitrogen from a source other than nitrates. 



These experiments are of the highest interest, and show 

 that much remains yet to be discovered with regard to the 

 conditions of bacterial life in soil, and its relation to the 

 growth of plants. 



Chemical Changes in Plant Cells. When a plant is burnt, 

 its organic constituents disappear, mainly as carbon dioxide, 

 C0 2 , nitrogen and water, H 2 ; its mineral constituents 

 remain behind in the ash. The growing plant builds itself 

 up again out of these products of its combustion ; the mineral 

 constituents and water it takes in through the roots, the 

 carbon and oxygen through the leaves, the nitrogen ulti- 

 mately being supplied from the sources already discussed. 

 All the complex physical and chemical processes involved 

 in building up a plant are controlled ultimately by the vital 

 energy of the plant cells, together with the energy of sun- 

 light. The correlation of all these processes is the task of 

 physiological botany, and a knowledge of this is obviously 

 inolispensable, if the plant is to be grown under the best con- 

 ditions and supplied with its right food. Enzyme chemistry 

 forms the foundation knowledge of physiological botany. It 

 is clearly necessary to have some understanding, in the 

 first place, of the primary chemical changes taking place in 

 specific cells, before establishing their general relations. 



The initial impulse to plant growth is to be found in the 

 potential biotic energy of the seed, or, more properly speaking, 

 of the embryo. This, of course, like all forms of ITe that we 



