PLANT TOXINS. A CAUS1<: OF INFERTILITY IN 

 SOILS: A SOUTH AFRICAN OBSERVATION. 



By Aktiuk Stead. B.Sc. 1'\C.S. 



(Plate 13.) 



The husbatulnian has ever been ])ersi.stent in liis l)eHef that 

 a croi) may poison the hmd for itself, so that when the same 

 crop is ojrown year after year on the same land the yields will 

 become .a^raduaily less, and finally dwindle to 'nothin<j. 



He has always beliexed in the rotation oif cro])s. by which 

 he ima.^ined decline in \ields was i^revented because the one 

 crop was not only not injured b}' the poisons of the other, but 

 actualh" had some j^ower to destroy the ]X)isons of the other. 



Ideas of this kind were crystallized l)y De Candolle in an 

 hypothesis enunciated at the be.s^inning nf last century and 

 favourablv commented on 1)\- Liebii^-. who wrote. al)out i<S4o: 



I )i.- Candolle sui)])0.m.'S tiiat tlu' rnots ni plants nnhine sniumo matter 

 .»f ever\- kind from tlie soil, and thus necessarily aliS(jrl) a number of 

 Hihstances which arc not adai)tcd to tlic pnrpnses of nutriticn. anfl must 

 ;ul)se(iuently be expelled by the roots and rettirned to the soil as excre- 

 ments. Xow as excrements cannot be assimilated liy tlie ))l;mt which 

 ejected them, tiie more of these matters which the soil contains, the more 

 im fertile must it be for plants of the same species. Tliese cxcrcmentitioiis 

 /natters ma\', however, still I)e capable of assimilaliim liy annthcr kind of 

 plant which would thus rcnioNc them from the soil and render it again 

 fertile for the lirst. And if the [)lants last tirown also expel sulistances 

 from their rc)ots which, can be approju'iati'd as fond by th^' former, they 

 ixill impi-o\c tile soil in two wny,." 



Certainly De Candolle's hypothesis must, in the li.nht of his 

 time, be considered as an inoenious and ])lausil)Ie attein];t at ex- 

 ])lainin;.)^ certain well-known facts cinniected with the rotation 

 of croj)s. For instance, in i'.nglatul cloxer cannot be jijrown two 

 \ears in successioti. while it leads to a considerable improve- 

 ment i;i the followini;- wheat cro)). Further, how may t)ne 

 explain the fact tliat certain ])lants i^rov,- better in association 

 than apart ? 



(1. C. Daubenv'^' inxestioated De Candolle's hypothesis by 

 i^rowing" t8 ditfereni crt)i)s under conditions of rotation as well 

 as otherwise. Me obtained declinitii"- yields iti all cases, while 

 there were matiy cases where the croi)s were orrown in rotation 

 in which the decline in yields was not so marked. 



De Candolle and Daui)env were ignorant of the effect of 

 tillafje o])erations in increasint;^ the output of the soil in ])lant 

 food and in conser\injj^ rainfall in tlie soil; neither did they 

 know that the benefit of clover to wheat consisted in the former 

 addinir intro_(^'en to the soil which, in bnii^iand. is the dominant 

 food for the wheat cro]); nor did the\' ai)pear to know that bv 

 .i.n'owiim' crops in rotation weeds, insect i)ests and diseases were 

 ke])t under better control. 



/'////. Trans. ( 1^15). 



