Chap, hi The absorption of Nitrogen 341 



1887, and by Frankland in 1888. These observers, together 

 with several others in later years, showed that many 

 organisms possess these reducing powers, and that the 

 ultimate products of their activity vary considerably. 



A third series of researches on the question of the nitro- 

 gen supply next claims our attention, one which was carried 

 on during the last twenty years of the century with con- 

 siderable success, and is probably about to lead to very 

 astonishing results in the future. It embraces questions 

 of symbiotic union between nitrifying organisms and the 

 higher green plants, a method by which at any rate certain 

 of the latter are able to appropriate atmospheric nitrogen. 

 The number of these known to show the symbiosis in 

 some form is continually increasing, and the most recent 

 investigations suggest that the method is very widespread in 

 the vegetable kingdom. 



In the early stages of this research, the question was 

 concerned chiefly, if not entirely, with the Leguminosae, 

 but other instances of similar symbiosis are continually 

 coming to light. 



Just before the opening years of our period, in 1858, 

 Lachmann had observed certain tubercular outgrowths on the 

 roots of certain leguminous plants and recognized that they 

 contained certain organisms much like bacteria. He associ- 

 ated them with an increased power to utilize the compounds 

 of nitrogen in the soil. In the next year Boussingault found 

 on growing lupins in rich garden soil, to which sand or quartz 

 had been added, that the soil gained considerably in its 

 content of combined nitrogen, in the forms of both nitrates 

 and ammonium compounds. No other plants than lupins 

 gave the same result. Boussingault put forward no very 

 clear pronouncement as to the cause of this gain ; he 

 denied, however, that it could have come from the air, 

 an opinion to which his experiments had led him in the 

 case of the normal green plant, as has already been pointed 



