No. 6. 



DEPARTMENT OF AGRICULTURE. 



373 



veiopment of Hie vetch took place when its roots were inoculated 

 with vetch organisms. In other words, a given legumen is most sus- 

 i;eptible to organisms of its own kind, and but feebly so to those of 

 another species. (See Fig. 10.) 



In the following table are given the results of these inoculations 

 by the above authors: 



Chemical Analysis of Plants. 



Inoculated With Pure Cultures From 



c 



!□ 

 o 



e 



3 

 n 



Roblnia dry substance— grams, 

 Robinia nitrogen — milligrams, 

 Acacia dry substance — grams. 

 Acacia nitrogen — milligrams. 

 Vicia dry substance — grams, . 

 Vicia nitrogen — milligrams, ... 



1.479 

 21.100 



1.817 

 19.700 



1.033 

 22.600 



The differences here noted can only be explained by a considera- 

 tion of the question of virulence of the organism, and of specific 

 adaptation. 



The pea tubercle organism by its growth in the pea root becomes 

 especially adapted to that host, and losses its adaptability to another 

 host. This, however, is a property' more or less elastic, and subject to 

 artificial modification. 



Thus Nobbe and Hiltner ^^ have shown that when peas were in- 

 oculated with cultures from bean tubercles, some tubercles would be 

 formed but the organism seemed to be without the power of assimi- 

 lating nitrogen; but if the same inoculation was continued upon a 

 second generation of plants, the bacteria became nearly as efficient 

 as those from the roots of the same genus. In other words, by pass- 

 ing pea tubercle bacteria successively through beans, the virulence 

 at first feeble was decidedly increased. 



The same authors have also shown that different cultures possess 

 different powers in producing tubercles. Thus if a plant already 

 possessing tubercles is inoculated with a culture of the same viru- 

 lence there is no increase in the number of tubercles, but when in- 

 oculated with a culture of a higher virulence there is an increase in 

 the number and size. The authors, therefore, hold that a plant may 



