Winifred E. Brenchlf.y 143 



generally recognised that up to a certain point it is profitable to give 

 growing crops plenty of room, l)ut that beyond this limit the total yield 

 is apt to fall off. 



Up to the present time most of our knowledge of competition has 

 been purely empirical and derived from observation and deduction 

 from plants grown under various cultural conditions not under strict 

 control. Consequently little information is available as to the relative 

 importance of the different factors that come into play when one plant 

 enters into competition with another, or of the interaction between 

 these factors. During the past five years a number of experiments have 

 been carried out at Rothamsted which aimed at isolating some of the 

 more important of these factors, and establishing their relationship to 

 the growth of the plant. A full discussion of the problem will not be 

 possible till far more data have been accumulated, but a preliminary 

 account will indicate the present position of the work. 



"Competition" of one plant with another is a very complex, not a 

 simple, phenomenon, and may be broadly analysed as follows: 



(1) Competition for food from the soil. 



(2) Competition for water. 



(3) Competition for light. 



(4) The possible harmful effect due to toxic excretions from the 

 roots, if such occur. 



The first three factors lend themselves to direct experiment; the 

 fourth is more difficult to demonstrate but the possibility of its existence 

 must be reckoned with in estimating results, 



1. GENERAL COMPETITION. 



(Including 1, 2, 3 above at the same time.) 



Pot cultures were carried out with mustard and barley, glazed 

 earthenware pots being used, each holding 10 or 20 kiios of soil. The 

 soil was obtained from an unmanured area on Great Knott Field, 

 Rothamsted, sifted, and mixed with 10 per cent, sand to lighten it 

 before use. No manures M^ere added in any case, so that the only mineral 

 food available for the plants w^as that present in the soil at the beginning 

 of the experiment. The seeds were sown in double quantity, but as 

 soon as germination was established the seedUngs were thinned out to 

 the correct number per pot. All seedlings thus removed and any weeds 

 that put in an appearance were laid on the surface of the soil in order 

 to avoid removing any of the original food supply and so altering the 



