5+- 



NATURE 



[January i6, 19 r 



lings in similar stages of growth were then trans- 

 ferred to the soil extracts — as was the case in the 

 writer's experiment above quoted — we still have in 

 the case of some varieties of plants this delayed ger- 

 mination period extended to a considerable length in 

 water cultures ; in other words, the extract from a 

 heated soil retards growth during a period of several 

 davs after the germination has actually taken place. 

 This appears to be connected In some way with the 

 formation of root hairs, the growth of which is often 

 entirely inhibited in water cultures. 



The' safest way of testing the effect of various soil 

 extracts on plant growth is to sow seeds in different 

 portions of one and the same soil (which should not 

 be too rich), and then, after germination, water with 

 the various soil extracts. 



Until some assurance is forthcoming that the neces- 

 sary precautions have been taken, the results of the 

 water cultures mentioned cannot be accepted. 



Finally, it will be difficult for any theory of soil 

 fcrtilitv that, like Dr. Russell's, claims that all, or 

 almost all, depends on bacterial activity to explain 



Fig. I.— Mai/e plants after growing for seven days in soil previously tre 

 heated ; 19-26 healed to 05' C. ; 27-34 heated to 17' 



away the injurious effect of any one plant on all it: 

 neighbours (see present writer in Journ. Agric. Sci. 

 vol iv., part iii.). F. Fletcher. 



Kvambu, British East Africa, November 27, 1912. 



Mr. Fletcher is under a misapprehension in attri- 

 buting to me a "theory" that "all, or almost all," 

 of soil fertility depends on bacterial activity. I 

 should certainly agree with Mr. Fletcher that any 

 such hypothesis, if it were advanced, would be much 

 too narrow to account for the facts. 



Soil fertility is not due to the operation of any one 

 factor, but of several. At least five conditions have to 

 be fulfilled by the soil if the plant is to make satis- 

 factory growth. There must be (i) adequate food 

 supply ; (2) proper water supply ; (3) suitable tempera- 

 ture ; (4) enough air for the roots ; (5) absence of 

 injurious substances or factors. Every one of these 

 conditions is essential ; any one that is unfulfilled sets 

 a limit to the growth of the plant, and therefore to 

 the fertility of the soil. 



I have discussed the interaction of these various 

 factors at some length in my book on " Soil Condi- 



NO. 2255, VOL. 90] 



tions and Plant Growth," and need now only refer to 

 the place of bacterial action in the scheme. 



Among the various nutrients required by the plant 

 are the nitrogen compounds. Nitrates are the com- 

 pounds usually obtained from the soil, but ammonium 

 salts also serve; there is evidence, however, that highly 

 complex compounds like the proteins, peptones, &c., 

 are of little value 'to the plant even when they are 

 soluble. Now the nitrogen compounds of the soil are 

 mainlv complex and insoluble, but they decompose 

 slowly to form ammonia, which then oxidises to 

 nitrates. 



It has been repeatedly demonstrated that when all 

 the other essential conditions are satisfied, an increase 

 in the supply of ammonium compounds or of nitrates 

 increases the amount of plant growth, i.e. of soil 

 fertility. An increased supply of ammonium salts and 

 nitrates may be brought about either by direct addi- 

 tion of these compounds or of substances easily con- 

 verted into them, or by increasing the rate at which 

 ammonia production takes place in the soil. 



The production of ammonia in the soil is largely 

 due to bacteria. When the condi- 

 tions are made more favourable to 

 bacterial action a marked increase 

 in activity sets in, accompanied by 

 an increased production of ammonia 

 and nitrate. A corresponding in- 

 crease in soil fertility follows. Par- 

 tial sterilisation of the soil leads to 

 marked increases in bacterial 

 numbers for reasons that Dr. 

 Hutchinson and I have discussed 

 elsewhere. The accompanying in- 

 crease in the amount of ammonia 

 produced is so closely connected 

 with that of the bacterial numbers 

 that no reasonable doubt can be 

 entertained as to its bacterial origin. 

 So much for the general relation- 

 ship of bacterial activity to soil 

 fertility. We can now turn to some 

 of the details raised by Mr. 

 Fletcher. He goes on to say that 

 if bacterial activity has anything to 

 do with soil fertility a completely 

 sterilised soil ought to be less fertile 

 than a partially sterilised soil. Un- 

 as follows:— u-is not fortunately no one has ever suc- 

 ceeded in carrying out this experi- 

 ment. When a soil is heated to 

 170° C, as in Mr. Fletcher's experiments, 

 or even to 120° C, as in some of ours, 

 it alters so completely tHat it can no longer 

 be compared in any sense with the unheated 

 soil. A considerable amount of decomposition takes 

 place, and much ammonium and other simple 

 soluble nitrogen compounds are formed. There 

 is no reason to suppose that it matters to the plant 

 whether the ammonium and other compounds are 

 formed by bacterial action or by any other process ; 

 the essential point is that they should be formed ; 

 whatever their origin, they serve as plant nutrients. 

 The increased gain in plant growth on such highly 

 heated soils can be largely attributed to this cause. 



The water-culture experiments, like the other ex- 

 periments made at Rothamsted, were carried out with 

 all the rare and precautions that we could command. 

 The obvious pitfalls mentioned by Mr. Fletcher were 

 avoided. The fact that our results differ from those 

 obtained by the United States Bureau of Soils implies 

 no contradiction at all ; they worked with " sour " 

 soils, and we worked with the entirely different "sick" 

 soils. We could find no evidence whatsoever of thp 

 presence of any toxin in our sick soils, or in oui^ 



