335 



The later experiments liave shown still further that the sterilized 

 legumes develoi) exactly iu proportion to the amount of nitrogen fur- 

 nishe<l in the soil and that if enough nitrogen is thus supplied the 

 plants grow luxuriantly and show a perfectly nornud develoi.iiient of 

 seeds. Of course a very abundaut supply of nitrogeu is necessary. 

 The cxperinieuts ha\'e showu that it nuist be applied iu a foiiii best 

 suited to the plants. It was found that ammonium nitrate is thr most 

 appropriate foiiii for supi>lyiiig leguminous ]»liiiits with nitrogen. Olher 

 coMijtonnds, such as cah-inm uitnite, do not agree so well with them. 

 Lupines fed with the latter compound become sickly and their devel«)p- 

 mcnt is imperfect. 



The statements just made refer to plants cultivated in pure sand by 

 Hellriegel's nu^thod. By this method healthy plants and especially a 

 normal development of seeds, can be obtained. This is illustrated bv 

 trials with lupine plants iu two pots of equal size, (uie c(.nt;iining 

 uMtural soil with bacteria, the other sterilized sau<l. Two plants were 

 grown in each pot. The weights of the plants produced (stems, leaAcs, 

 and fruit, but not the roots) were, (1) in a natural soil with syml»iosis, 

 41. OS grams of dry .substance, of which the seeds were 25.7 jxt cent; 

 (2) iu sterilized sand with ammonium nitrate, 40.71> grams of diy 

 substance, of which 26.7 per cent were seeds. 



A unmber of experiments were made in jiots with natural soil 

 Whenever the soil was not sterilized the leguminous plants jiad 

 tubercles on the roots and a notable acquisition of nitrogen was found. 

 When they were kept sterile the plants grew only in proportion to the 

 nitrogen contained in the soil, l)utwlien nitrogenous fertilizing material 

 was added to the sterilized cultures, the amount of plant growth 

 increased in proportion to the amount of nitrogen supplied. The roots 

 had no tubercles and there was no evidence of acquisition of atmos- 

 l>heric nitrogen. In this way the same ininciple is found to ap]>ly to 

 culture in soil containing humus, as in pure sand. The view of Fiaidc 

 who claims especial effects in humus soils, is thus refuted. 



Besides these pot experinu'uts a series of field trials were made with 

 lupine and serradellato study the effects of inoculating with bacteria. 



The experiments at Bernburg bring out the fact (which has been 

 confirmed by other observatnms, e. g., those of Nobbc, described in the 

 present number of the Eecord, page 336) that different leguminous 

 plants do not avail themselves of the same kind of bacteria, but rather 

 that a given species may require a special form for the syndnosis l»y 

 which the free nitrogen is fixed. Thus, for instance, the root tubercle 

 bacteria of peas do not produce tubercles in lupines and serradella, 

 and therefore do not enable them to acquire nitrogen i'rom the air. 

 Now these different forms of bacteria are not found in all arable soils. 

 Thus the cultivated soils in the vicinity of Bernburg contain an abund- 

 ance of pea bacteria, but none or very few of the lui>ine bacteria, because 

 lupines have never been cultivated in this region. On the other hand, 



