304 



We see from these data that commonly only a small part of the sown germs 

 comes to growth. Whether perhaps the water itself has a deadly influence on some 

 individuals, or that their death is caused by their passing on the solid medium, could 

 not yet be made out by experiment. Thus, although there be ground to allow that 

 more germs occur in the soil used than are found, the possibility exists that by 

 continued investigation the experiment may be made so as to exclude that source 

 of error. 



But in spite of the uncertainty of the method the following result could be stated. 



By sowing a small quantity, for instance less than - - gram of garden soil on calcium- 



malate-kaliumphosphate i% agar, after 24 hours at 30 C. commonly no Azotobacter 

 is observed, but a moderate number of moist colonies of about i mm. in diameter, 

 first draw attention by their extension and prove to consist of different varieties of 

 Bacillus megatherium, containing many spores. They dont cause any considerable 

 oxidation of the malate and as the colonies no more grow after the second day, they 

 evidently develop at the expense of the traces of nitrogen compounds which at first 

 are present in the plates. After the second day a great number of Streptothrix alba 

 appear. This microbe is so common in all the examined samples of soil that there 

 can exist no doubt as to its either favourable or pernicious influence on the fertility; 

 but the nature of this influence is as yet wholly unknown. 



In a still later stadium the surface of the plate becomes covered with numerous 

 relatively small colonies of bacteria, among which some species immediately draw 

 attention by their extension and commonness. 



The oxidation of the malate by all these microbes is slight, so that even after 

 weeks the plates contain but little calcium carbonate, which seems almost entirely 

 produced by the said larger colonies and by Streptothrix. All these species seem not 

 to oxidise at all, or perhaps it is more accurate to say, not to oxidise any more after 

 the last traces of fixed nitrogen have been assimilated. As to Streptothrix, from its 

 relatively vigorous oxidising power it follows by no means that this should be 

 associated with fixation of nitrogen; this species surely does not possess that faculty. 

 If for the experiment soil is used shaken from the roots of garden-plants, which are 

 no Papilionaceae, the result is fairly the same; perhaps the number of the above men- 

 tioned oxidising forms is more numerous, but this is still doubtful. 



Otherwise, however, is the result when the soil is examined which adheres to 

 the roots of clover, pease, and beans when these plants are cautiously dug up. When 

 the soil adhering to such roots is rubbed fine and after dilution in water sown on a 

 malate plate we find, after a period of 2 days at 30 C., first that the said oxidising 

 colonies have very abundantly developed. But, besides, among these colonies much 

 larger ones are distributed, which oxidise much more vigorously and prove to belong- 

 to Azotobacter, which shows that a distinct relation exists between the distribution 

 of this genus and the said Papilionaceae. Whether this relation will appear to be 

 universal and what may be its signification, further experiments have to decide. 



