iriH SOIL AND ITS EFFECT ON PLANT GROWTH. 13 



sted fields during the years 1904-1917. The object of the experi- 

 ments was to test the residual value of a number of manures, 

 including farmyard manure, shoddy, guano, rape cake, super- 

 phosphate, bone meal and basic slag. No regular rotation was 

 adopted, but clover was avoided. After the experiment had 

 gone on for twelve years it was decided to make a change and to 

 introduce clover after mangolds. It was shown that none of the 

 manures except farmyard manure had any marked effect in 

 increasing the clover crop. Rape cake, guano and shoddy had 

 no effect at all, even when applied to the preceding wheat crop ; 

 superphosphate, bone meal and basic slag had a small effect but 

 not definite. Farmyard manure, however, had a striking effect, 

 the clover crop on these plots being approximately half as heavy 

 again as on the remainder. Subsequent field trials have confirmed 

 these results, and show very clearly the markedly beneficial 

 effects of the application of farmyard manure to clover, particu- 

 larly in conjunction with lime. An investigation has therefore 

 been commenced to find out the reason. 



It should be explained at this stage that leguminous plants 

 such as clover are especially valuable to the farmer because 

 their roots form a " breeding ground" for certain kinds of 

 bacteria which are present in the soil and which have the power 

 of extracting nitrogen from the air. These bacteria live and 

 multiply in nodules on the roots of the clover plants, so that 

 when the crop is ploughed in, it supplies the soil with a quantity 

 of valuable nitrogen which can be used by the succeeding crop. 

 The type of bacterium which can live on clover roots, however, 

 is present in large numbers in the Rothamsted soil, so for the 

 purpose of accurate experimental work it was necessary to choose 

 another leguminous plant — the soy bean — whose roots are not 

 acceptable to the " clover " type of bacteria, but to another type 

 which is absent from the Rothamsted soil. By using the soy 

 bean greater control can be exercised by adding or withholding the 

 bacteria which will form the " nodules " on the roots of the plant. 



The laboratory experiments were started in 1920 and have 

 thus been carried on for two years. Seven series of pots 

 growing soy beans have been taken, each series being treated 

 differently, as follows : — (i) i lb. farmyard manure : (2) | lb. 

 farmyard manure ; (3) liquid extract from | lb. farmyard 

 manure ; (4) mineral salts giving the same amount of plant 

 food as I lb. farmyard manure ; (5) no manure ; (6) h lb. fresh 

 straw ; (7) | lb. rotted straw. The experiment was carried out 

 both in the presence of " nodule " bacteria and in their absence. 

 For each series, information has been obtained regarding the 

 number of nodules formed, the total dry weight of the plants, 



