1 70 CONTRIBUTIONS TO OUR KNOWLEDGE OF SOIL-FERTILITY, xvi., 



The behaviour of Pen. cladosporioiaes was different in tViis ex- 

 periment from that in the presence of an excess of chalk. It did 

 not decompose the organic matter so quick!}', but the rise in the 

 dilution-curve was more pronounced. This suggested the idea 

 that, if the nutritive effect could be kept down, the curve might 

 be more accentuated Use was made of Rhizobium for growing 

 with the Penicillium because, when the former was tested at the 

 same time, viz., on the tenth day, the nutritive ratio was lower 

 than either the latter or the control-test. The double growth 

 kept down the ratio, but there was no elevation of the dilution- 

 curve. The growing Penicillium brought about an alteration in 

 the reaction of the extract. The reaction of the control-test 

 seems to indicate, that the bacteria and moulds cause the medium 

 to furnish acid extracts which, in time, may become alkaline. 



The effect of using different amounts of water in preparing the 

 extract was shown in the case of the mould infected with Fusa- 

 rium species. Two tests were extracted at the same time, one 

 with 500 c.c, the other with 250 c.c. of water. The stronger 

 extract gave the higher nutritive ratio, as 204 : 42, while the 

 weaker solution showed a tendency to produce a rise in the 

 dilution-curve. 



On the whole, the neutralised leaf-mould did not come up to 

 expectations in serving as a suitable medium for the demonstra- 

 tion of toxin-production by the selected micro-organisms. 



When an extract is diluted and subsequently sown with 

 bacteria, it will give a count in proportion to the amount of 

 dilution. But if the extract is acid, the dilution will be less 

 acid, and the count will not be in proportion, but will be more 

 or less according to the effect of the change in the acidity upon 

 the growing bacterium. It is difficult to say just what this 

 difference from the normal would be, especially with such weakly 

 acid extracts as in Expt. xix. The neutralisation of an extract 

 so strongly acid as -f TO is well seen in Expt. xvi., in which the 

 dilution to one-fifth raised the count from 150 to 300-fold. A 

 set of experiments were made to obtain some definite informa- 

 tion regarding the influence of slight variations in the reaction. 



