JO Feb., 1910.] Tests with Cultures of Roo't-tubcrclc Bacteria. 103 



The central line of plots, and more especially the clover plots, suffered 

 through the lack of drainage during the very wet winter season, and the 

 plots were unfortunatel\ not weeded as evenly and simultaneously as they 

 might have been, which is of considerable importance in maintaining 

 uniform conditions. In regard to the previous history of the plot, no 

 legumes had been grown in the ground for years, but only roots and fodder 

 crops. 



Last season, howe\er, some manure tests had been made, including 

 some complete sets. No organic nitrogen, however, had been applied 

 except to section D. This fact was not known at the time the experiments 

 were begun, and it is responsible for a certain amount of fluctuation in 

 the results from the different plots. To some extent, however, the dis- 

 crepancies are removed by the fact that each experiment was done in 

 duplicate, and that we may take a sterilized inoculation, where the culture 

 had been boiled for one hour before the seeds were moistened, as being 

 practically the same as not inoculating at all. On this basis', the in- 

 oculated plots of peas averaged 312 lbs., the uninoculated, 252 lbs., that is 

 a difference within a 20 per cent, limit of error. As a matter of fact, the 

 apparent slight benefit derived by the inoculated plots, is due to the 

 partial failure owing to climatic conditions of the central line of plots. 



Thus the bottom plot at D gave a greater yield than the inoculated 

 one, and if the comparison is made between the inoculated plots and 

 those where the culture was sterilized, the results agree within a 10 per 

 cent, limit of error. In the same way, comparing the peas with and 

 without lime, although the lime appeared to favour growth (780 lbs. 

 without to 844 lbs., with lime) the difference is less than 10 per cent, 

 and the biggest yield of all was given on one of the unlimed plots, 

 324 ibs. 



In the case of clover, the smallest vield of all was given by one of the 

 inoculated plots in the central row while the next highest was' given by 

 one of the inoculated on the outside row, and the highest of all by an 

 uninoculated plot (59 lbs. 7 ozs.). The average for the inoculated plots 

 was 41 lbs. 10 ozs., and for the sterilized inoculation plots 44 lbs. 6 ozs., 

 which are agriculturally identical yields. This is because one of the 

 inoculated plots was in the bad central row, and one in the good outside 

 row, so equalizing the conditions, whereas with the peas, both the inoculated 

 plots were on the outside. 



The root systems of selected plants were lifted from all the plots and 

 ■compared. The tubercles on the peas were especially abundant and 

 large, those on the clovers were small, few in number, shrivelling and 

 turning brown at the time the examination was made. No signs of any 

 beneficial effects of inoculation upon the development of root-tubercles 

 could be observed in any of the plots. 



The net result of all these experiments is therefore to show — - 



{a) Infection is possible in a water culture. 



{b) Root -tubercle bacteria must be able to exist in the soil for at least 

 ten years, for no legumes had grown in the University plot, and no 

 manure had been added to it for at least that period of time. 



{c) In soils' containing these bacteria, even if they are not abundant, 

 no appreciable benefit will be derived by inoculating the seed or soil with 

 cultures of root-tubercle bacteria. The number of bacteria .so added is 

 trivial as compared with those already present in the soil. 



{d) On the Burnley plots', the distribution of the water when in excess 

 influenced the subsequent growth of the plants more than an\- other factor. 



