September 22, 1921] 



NATURE 



117 



manufacturini^ facilities are, perhaps. ade- 

 quate for present demands, but it is certain 

 that much more fertiliser could be used, and 

 that as farming improves the demand will 

 increase. 



Progressive farmers have long passed the stage 

 when it was necessary to demonstrate that arti- 

 ficial manures increase crop production ; the posi- 

 tion now is the much more difficult one of deciding 

 how much money it is wise to spend on fertilisers. 

 The old view was that the crop yield was propor- 

 tional to the manurial dressing — i.e. that the more 

 the manure the bigger the crop. Lawes and 

 Gilbert showed this was not altogether correct, 

 and that the yield fell off "after a certain sized 

 dressing was reached; this relationship is ex- 

 pressed by a straight line which ultimately be- 

 comes a curve. A later view set up by Mitscher- 

 lich was that the effect of the manure is propor- 



Increased 



>-ield of grain. 



Bushels per acre. 



Increased 

 >ield of straw. 

 Cwts. per acre. 



Old 



Idea 



Effeci" of FeTh user 

 Qw Cio\t Yield. 



Later 

 Idea. 



Presc nf 

 lcU&. 



rroportional, 



Tri€n Diminisniiw 



Return. 

 ^sTRAicHT line) 



Con ti nuouslj 

 fall i MO-. 



Small, 



Tnen mcreasin 



Tnen Decreasing. 



<L«CAR|THMIC CUHVt) (jiCMOIP.aaiun'tCATALrTIC CMV(> 



Fig. I.- 



Zurves showing relationship between crop yield (plotted on vertical axis) and 

 quantity of fertiliser used (plotted on horizontal axis). 



tional to the decrement from the maximum obtain- 

 able ; that therefore the first dose of manure has 

 a large effect ; but that further doses have pro- 

 gressively less action. This relationship is ex- 

 pressed by a logarithmic curve. The present view 

 is that the effect is at first small ; then it in- 

 creases and then decreases; this relationship is 

 expressible by a curve resembling that for auto- 

 catalysis. The important practical consequence 

 is thaf moderate dressings are more profitable 

 than small ones, but they are also more profitable 

 than much larger ones (Fig. i). There is no 

 difficulty about the general rule ; the difficulty 

 arises when one tries to define a moderate dress- 

 ing. The problem is further complicated by the 

 fact that the effect of the dressing is greatly in- 

 fluenced by the time when it is put on to the 

 land. In our own case the results have been as 

 follows : — 



NO. 2708, VOL. I08I 



Date of application of manure Feb. lo Mar. 6 May lo Feb. lo Mar. 6 May lo 

 Single Dressing . . Nil. 0-9 2-7 2-7 6-9 9-4 



Double Dressing .. 70 — 3-7 11-7 — 12-7 



This experiment ought to be repeated in many 

 districts, for it is by no means certain that farmers 

 generally are using the most profitable quantities 

 of fertiliser at the most effective time. It is, 

 however, necessary to take into account some- 

 thing more than the quantity and the time of 

 application of the fertiliser. It is essential also 

 to have a suitable mixture. In the old days this 

 question was thought to be fairly simple. 

 Chemists used to think that if they knew the 

 composition of the ash of plants they would know 

 what manure to use; it should supply all the ash 

 constituents in the quantities present in the plant. 

 This is now known to be wrong; the composition 

 of the ash affords no guidance to manurial 

 requirements, as was, indeed, shown by 

 Lawes and Gilbert in 1847. The distin- 

 guished French chemist, Georges \'ille, 

 emphasised the fact that onlv properlv 

 conducted field trials would ever settle the 

 question. Vast numbers of such experi- 

 ments have been made, and thev show- 

 that the problem is more complex than 

 \'ille thought. It is now known that no 

 single formula expresses the fertiliser 

 needs of a crop ; every district, almost 

 every farm, has its own special require- 

 ments. 



Still further difficulty is introduced by 

 the fact that the various artificial fertilisers 

 not only increase crop yields, but also 

 influence the composition and habit of 

 growth of the crop. Nitrogenous 

 manures tend to a vegetative growth of 

 large, deep-green leaves which are some- 

 what liable to be attacked by fungoid 

 pests. Phosphates improve root develop- 

 ment, and are therefore of special value 

 for swedes and turnips ; they also hasten 

 ripening of grain, and are therefore 

 particularly useful in late districts ; they increase 

 the feeding value of crops, and are therefore use- 

 ful for fodder crops; and they have a remarkable 

 effect on the development of clover, which is not 

 yet fully understood, but which has revolutionised 

 the treatment of pastures in this country. Potassic 

 fertilisers improve the vigour of the plant and 

 ' increase its power to resist fungus attacks. These 

 and other special properties of fertilisers are now 

 well established, and advantage is taken of them 

 in drawing up fertiliser schemes to suit the special 

 requirements of each farm. 



It has already been pointed out that this work 

 on artificial fertilisers arose out of Lawes and 

 j Gilbert's discovery that the wheat crop of 1843 

 ' grew just as well when supplied with the ash con- 

 I stituents phis combined nitrogen as when supplied 

 ! with farmyard manure. They repeated the ex- 

 j periment year after year; periodically the results 



