Relations of Higher Plants to the Formation of Nitrates 87 



to such an extent that the nitrogen in the crop plus that in the drainage 

 water was less than the nitrogen in the drainage water from an unplanted 

 lysimeter containing a similar soil. These results are in part confirmed 

 by experiments conducted by the writers with somewhat similar appa- 

 ratus. The writers have had only two years in which to use this set of 

 tanks, and hence the data are probably not entirely reliable. 



Experiments were conducted in large concrete tanks filled with soil. 

 The tanks are four feet two inches square and four feet deep and each tank 

 holds about three and one half tons of soil. The soil used was from the 

 experiment field in which most of the other experiments with nitrate for- 

 mation were conducted. The tank soil receives the natural rainfall and 

 no other water supph'. The drainage water is all collected and measured 

 and samples are taken for analj'sis. In 1910 four tanks were planted to 

 maize, two were planted to oats, and three remained unplanted. In 1911 

 four tanks were in oats, two were in grass, and two were bare of vegetation. 

 The unplanted tanks were cultivated in 1910, and in 1911 they were 

 scraped in order to remove weeds mthout stirring. 



In Table 34 are given the quantities of nitrogen in the drainage water, 

 and in the crop, including the estimated quantities in the roots, for the 

 periods already stated. 



The results do not agree in every respect mth those obtained by Leather 

 and by Deherain. Both the tanks pla,nted with maize and those planted 

 with oats contained a larger quantity of nitrogen in the crop and the 

 drainage water combined than the unplanted tanks contained in the 

 drainage water. The grass-planted tanks, like those of the experimenters 

 just cited, contained notably less. In fact, less than two thirds as much 

 nitrogen was found in the combined crops and drainage from the grass 

 tanks as was found in drainage from the bare tanks. 



The quantity of nitrogen removed by the crop and the drainage water 

 combined is in the same order with respect to the kind of plant grown as is 

 the nitrate content of the soil, as shown by the plat experiments. Thus, 

 the nitrogen removed from the maize plats was greatest, that from the oats 

 plats was next, and that from the grass plats was least; and the nitrate 

 content of the maize soil is higher during the grooving season than is that 

 of oats, and the latter soil has uniformly a higher nitrate content than has 

 grass land. These results are shown graphically in Diagram XXV. 



