150 CONTRIBUTIONS TO OUR KNOWLEDGE OF SOIL-FERTILITY, xvi., 



Experiment ix. 



There are indications of the formation of toxin in some of these 

 tests, although it must be said that they are not very pronounced. 

 Test 1 shows a narrow extract/water ratio, pointing either to the 

 possible absence of nutrients in the extract or to the presence of 

 toxins. The increases obtained in the boiled and in the diluted 

 extract lead one to believe that toxins were present. Test 2 had 

 undoubtedly been incubated for too long, but there is a suspicion 

 that the extract is of the same nature as test 1. Test No. 4 is 

 irregular, and, so far as we can judge, gives us no information. 



In these tests, the growing organism had undoubtedly removed 

 nutrients from solution, and, by utilising them, had prevented 

 their appearance in the porcelain filtered extract. This was 

 shown by an extension of test 1 , in which the raw, uninoculated, 

 filtered hay-infusion gave 666,800 cells, and the same, when 

 boiled, gave 379,600 cells as against 100 of the unboiled bacterial 

 extract. 



A mixed culture of bacteria, existing as an actively nitrifying 

 suspension of soil-bacteria, was seeded into a solution of 0*1% 

 dextrose in 0-1% hay-infusion. The suspension was incubated 

 at 15° for 7 days. On preparing and testing the extract, it was 



