38 EXPERIMENTS UPON WHEAT 



all. It has been already pointed out that the Rothamsted soil 

 is by no means exceptionally rich, how then can this continued 

 production of crop without manure be accounted for ? It is 

 estimated that the average crop on this plot has removed about 

 17 Ib. of nitrogen, 9 Ib. of phosphoric acid, and 14 Ib. of 

 potash per acre per annum. In the drainage water there is 

 also a further loss of nitrogen, which has been estimated at 10 

 Ib. per acre per annum ; some nitrogen is also removed in 

 weeds. Per contra, the rain brings about 5 Ib. of nitrogen each 

 year, and the seed supplies perhaps 2 Ib., thus leaving a nett 

 annual loss of nitrogen of at least 20 Ib. per acre. The 

 analyses of the soil taken in 1865, 1881, and 1893, show that 

 there is a steady diminution in the amount of combined 

 nitrogen present in the soil ; but since in 1893 the proportion 

 present was 0'099, or rather more than 2500 Ib. per acre in the 

 top 9 inches of soil, there is still an enormous reserve un- 

 touched. There may also be hitherto unrecognised gains of 

 nitrogen from the atmosphere. For example, the Black 

 Medick is a common weed on this plot, and like other legu- 

 minous plants fixes some nitrogen from the atmosphere, part 

 of which will be left behind in the soil when the roots decay. 

 Soil bacteria are also known which are capable of fixing 

 nitrogen independently of the higher plants; but until the 

 analyses of the soil have been repeated after another long 

 interval it is not possible to say whether such recuperative 

 agencies have any practical effect, or whether the crop is still 

 being grown out of the original resources of the soil. As regards 

 potash and phosphoric acid, there can be no external sources 

 of recovery, but the reserves are very great, amounting in 1893 

 to about 3000 Ib. of phosphoric acid and as much as 50,000 

 Ib. of potash per acre in the top 9 inches of soil, though of 

 course the greater part of the latter would only become avail- 

 able for the plant very slowly. 



On Plot 2 there has been a yearly dressing of 14 tons of 

 farmyard manure, and though the composition of the dung is 

 so far variable that it is impossible to say exactly what 



