66 Martin J. Frucha 



In this test a knowTi number of organisms was introduced into each 

 flowerpot at the time of inoculation, in order to trace the relation between 

 the number of nodules developed and the number of organisms used for 

 inoculation. It was hoped that m this way some means might be found 

 of measuring not only the loss, but also the degree, of infecting power. 

 From the data m table 15 it is seen that absolutely no relation exists 

 between the nmnber of organisms introduced into the flowerpots and the 

 number of nodules that developed, so that the only measure of mfecting 

 power is the presence or the absence of nodules. This, however, does not 

 measure the degree oi mfecting power, but only its presence or absence. 



The average number of nodules per plant produced by each culture, as 

 shown in table 15, are rearranged in numerical order in table 16. It is 

 found on comparing tables 13 and 16 that the results of the third test of 

 infecting power show a general agreement with the results of the second test. 



If it is considered that the presence of three or more nodules per plant 

 indicates that the culture was eSicient, the following cultures are seen to 

 have lost their efficiency: 



Soy bean hay 

 Soy bean roots 

 Canada field pea hay 

 Canada field pea roots 

 Medium 334 + 10 per cent cane sugar 

 Medium 334 + 40 per cent cane sugar 

 Medium 334 + 0.2 per cent phloroglucin 

 Medium 334 + 1 per cent salicin 

 Medium 334 + 0.5 per cent amygdalin 

 Medium 334 + 0.5 per cent resorcin 

 Medium 334 -}- 0.5 per cent potassium oxalate 

 Medium 334 + 5 per cent Witte's peptone 



The cultures that had any living organisms at the time of inoculation 

 produced nodules practically m all cases m the three tests. The cultures m 

 which no living organisms were found produced no nodules, or only a few 

 unevenly distributed. It must be remembered that a certam amount of 

 contaminationmay occur, and that the method of determining the presence 

 of living organisms by plate cultures is not absolutely accurate. The cul- 

 tures in this test were ten weeks older than those in the second test. The 

 results show that the nodule-bacteria cultures can be kept for at least 

 twenty weeks, and the bacteria will still be efficient m producing nodules. 



