582 



NATURE 



[January 20, 19 16 



of infertility has to be studied the first question to settle 

 is : What is the limiting factor? And the next : 'How 

 •can this limiting factor be put out of action? As a 

 rule the limiting factor is one of the following : — 



Limiting factor. Put out of action by : 



Wetness Drainage, liming 



Dryness Irrigation, suitable cultivations 



Addition of organic matter 

 Lack of temperature Drainage and cultivation 

 Sourness or acidity Liming or chalking 



The removal, although simple in principle, may be 

 very difficult in practice : it has often proved to be 

 the rock on which many beautiful schemes for increas- 

 ing food production have been wrecked. 



We have seen that, broadly speaking, three general 

 principles of crop production can be laid down : — 



(i) The plant must have a sufficient supply of all 

 necessary nutrients, especially of nitrogen, potassium, 

 iind phosphorus. 



(2) The biochemical decompositions in the soil must 

 proceed smoothly and quickly. 



(3) All the requirements of the plant must be satis- 

 fied. Any one left unsatisfied constitutes a limiting 

 factor preventing further growth. Increases in any 

 one factor give increases in growth until some- 

 thing else proves insufficient and becomes a limiting 

 factor. 



We go back, then, to our three established prin- 

 ciples. Each of these can be recognised broadly in 

 every case of crop production, but considerable difficul- 

 ties arise when one tries to develop any of them ; there 

 are so many factors involved and their interaction is so 

 complex. I can best illustrate this by taking one of 

 the factors in some detail, and I will choose one that 

 has received very much attention from chemists, 

 namely, the supply of phosphates. 



Phosphates are indispensable for plant growth, and 

 well conducted physiological experiments in sand have 

 shown a simple connection expressible by a mathe- 

 matical equation between the amount of phosphate sup- 

 plied and the amount of growth. But such simple 

 results are never attained in soil. To begin with, there 

 is always some phosphate alreadv present. At first 

 sight it looks easy enough to take account of this, 

 and simply add it on as a constant in the equation. 

 It has proved almost impossible, however, to give any 

 precise value to the amount of phosphate in the soil 

 that is of any use to the plant. Ville showed years 

 ago that the amounts revealed by chemical analysis 

 are far beyond anything the plant can ever get, and 

 he rather gloomilv concluded that "chemistrv is power- 

 less to throw light on the chemical properties of the 

 soil." One could scarcely expect chemists to acquiesce 

 inthat view, nor did they, 'instead of using strong 

 acids, they used dilute acids ; several were suggested, 

 and by a happy inspiration Bernard Dyer selected 

 I per cent, citric acid as being the most suitable; 

 although that was twentv-one 3-ears ago, i per cent, 

 citric acid still holds the field in' this countrv. 



The part extracted by dilute acids was' called the 

 "available" portion to distinguish it from the "un- 

 available." The new method at once proved very help- 

 ful ; difficulties, however, began to arise. It was found 

 Impossible to assign any definite value to the amount 

 of available phosphate present. Variations in the con- 

 ditions of the experiment gave wholly different values 

 for the amount of "available" phosphate, whilst in 

 the case of nitric acid the longer the acid acted the less 

 phosphoric oxide (PjO,) was extracted. 



Now that gave the clue to the problem. It is obvious 



that there must be two actions going on : a direct 



solvent action and a reverse action, resulting in the 



withdrawal from the solution of the dissolved phos- 



NO. 2412, VOL. 96] 



phoric oxide. The direct solvent action was found ; 

 be much the same for all dilute acids. 



The reverse action proved to be the ordinary adsorp- 

 tion isotherm, similar in type to that obtained with 

 charcoal and dilute acids. The constants are 

 not the same for the different acids, and from 

 these curves it is possible to go back and explain 

 the apparently erratic action of the different acids on 

 the soil. 



Thus it appears that when phosphate is added to the 

 soil for the purpose of increasing the grovyth of a 

 crop it does not simply stop in the soil, waiting for 

 the plant to take it up. It reacts with the soil ; it is 

 adsorbed, and the amount available for the plant at 

 any time depends on the adsorption relationships. 

 There is, in short, a competition between the plant and 

 the soil for the phosphate. The curves for clay and 

 sandy soils show that adsorption is greater for clay 

 than" for sand; in other words, the clay competes for 

 the phosphate more vigorously than does the sand. 

 An amount, therefore, which is sufficient for the plant 

 growing in a sandy soil proves inadequate on a clay 

 soil. This has thrown light on an interesting problem 

 in manuring, for it has long been known that clay 

 soils stood in more need of phosphatic manures than 

 sands. The field results bring out this fact : the yield 

 of barley on the heavy Rothamsted soil falls when 

 phosphates are omitted', but it does not react- nearly 

 so quickly on the W^oburn sand. 



It seerns a far cry from the logarithmic curve ex- 

 pressing an adsorption isotherm to the management 

 of barlev and turnips, but the connection is reallv 

 simple and direct. 



This, however, does not settle the matter. Tlv 

 plant is a living thing, and consequently its require 

 ments are not rigidly constant, but vary with t! 

 conditions. There are very good grounds for suppl- 

 ing that the plant actually requires more phosphate on 

 a clay soil than on a sand. The effects produced by 

 phosphates are promotion of early growth, root de- 

 velopment, and early ripening ; they are specially valu- 

 able on clay soils, in wet regions, and for shallow root- 

 ing and quick growing plants, for example, swed* - 

 and turnips. 



It is possible that some simple connection underlii - 

 all these, but no one has yet discovered it. 



Again, seeing that the need varies with the condi- 

 tions, it is clear that if the conditions are altered the 

 needs may change. When, for instance, a dressing of 

 farmyard manure is applied, some of the properties of 

 the soil are altered ; it becomes more porous and mor<' 

 retentive of water, and phosphates may behave differ- 

 entlv from what they did before. That is well shown 

 on the Saxmundham plots. 



It is unnecessary to go any further. The point I 

 want to bring out is that the simple and incontrovert- 

 ible statement that phosphates increase plant growth 

 proves very complex when aoplied in practice. So it 

 is with the other factors. They can be disentangled 

 and investigated, but it is not yet possible to put them 

 together again and predict the resultant. One cannot 

 set out from first principles and reconstruct the norm:il 

 case of crop production ; the factors are too numerous 

 and too complex. 



Yet something has got to be done. The technical 

 chemist has the advantage over his colleague in the 

 purely scientific laboratory that he cannot shelve an 

 inconvenient problem. A method has been evolved, 

 an empirical method, which, whilst not very rigid, has 

 at least the merit that it works. It consists in going 

 into the field and finding out the actual agricultural 

 properties of the soil by observations, inquiries, and 

 direct field experiments ; these have to be repeated for 

 two or three years because the first results may only 

 have been a trick of the weather, but if the same 



