Feb. 1897.] BOTANICAL SOCIETY OF EDINBURGH 31 



will have rendered the water cloudy, will be found many of 

 the nodule-making forms. The bacteria penetrate a root 

 hair or a non-cuticularised epidermal cell (the young parts 

 are the seat of infection, the older parts of the root not 

 allowing an entrance), where they multiply at the expense 

 of the cell contents. The irritation set up in the cortex 

 cells, in which the bacteria ultimately find themselves, 

 gives rise to repeated cell division, and the wart is formed. 

 Finally, these bacteria take on a forked or Y-shaped 

 appearance, when they are known as bacteroids. The full 

 significance of these bacteroids will appear later, sufficient 

 here to say that their contents are finally absorbed by the 

 host plant. A certain number of bacteria remain behind, 

 which, on the breaking down of the nodules, are set free 

 again in the soil, where they may proceed to new infection. 



Frank had previously noticed that nodules did not 

 appear on leguminous plants grown in sterilised soil, but 

 the proof that the Leguminosaj are not able to make use 

 of the free nitrogen in absence of nodules, but only when 

 nodules are present in their roots, was given by Hellriegel 

 and Wilfarth, after a series of very accurate recorded 

 experiments. Their experiments showed that peas and 

 lupins grown in soil sterilised, but with all the necessary 

 plant foods added save nitrogen, died off without any 

 nodule formation as soon as the original store of nitrogen 

 present in the seeds had been exhausted. Peas and lupins, 

 on the other hand, sown under the same conditions, save 

 that a little soil was added from a field where peas and 

 lupins had been growing the year before (in which soil 

 it was reasonable to expect the infecting bacteria to be 

 present), showed a large formation of nodules and a most 

 excellent growth. 



Many workers, British and foreign, have since verified 

 these results, but Professor Nobbe and Dr. Hilltner, of 

 Tharand, have taken a step in advance by being the first 

 to induce nodule formation by inoculation with j)ure 

 cultures. Here is the method of making such pure 

 cultures : Take a fresh nodule and wash it carefully. 

 After drying it in blotting paper, drop it for a moment 

 into corrosive sublimate, to kill any bacteria on the 

 surface. Next wash it with absolute alcohol. Having 



