TOXIC ACTION OF ONE CROP ON ANOTHER 297 



the reverse ; but for any strict comparison it would be necessary 

 to have the same amount of surface growth in the various cases : 

 this might, perhaps, be accomplished, but it would necessitate 

 numerous series of experiments, which, with these heavy and 

 bulky pots, would require many seasons for their completion. 



For these same reasons conclusive evidence has not yet been 

 obtained on another point of considerable importance, namely, 

 whether a plant when grown as a surface crop will affect a plant 

 of the same nature to a greater or lesser extent than it affects a 

 plant of a different nature. That it does affect one of the same 

 nature as itself, there can be no doubt, and the evidence as it 

 stands indicates that it affects it even more than plants of a 

 different nature. Thus 



Relative growth of a plant when acted on by other plants. 



A. Of the same sort. B. Of a different sort. 



Grass . (24 to 31=) 38-1 (29 to 85 =) 69 

 Mustard . (i to 40 =) 21 



Clover i6>27 (3 to 77 =) 55 ^55 



Tobacco ... 36 ... 

 Tomatoes . . . 33' 



In these experiments it was often evident that the influence of 

 the surface crop was effective, not only in reducing the total 

 growth of the plant, but in delaying its development. This was 

 not always the case, but when it did occur, it introduced some 

 difficulty in making a comparison between the growth of the 

 plants with and without the surface crop, for maturity was not 

 reached at the same time. By leaving the plants with the 

 surface crop longer to develop, greater errors, however, were intro- 

 duced, for a considerable increase in weight occurs in a plant 

 after it has apparently completed its functions and ripened its 

 fruit. This led to an erroneous conclusion that plants under the 

 influence of a surface crop would, before the end of the season, 

 more than recover the ground lost, and would outstrip plants 

 without any surface crop (XIV, Append. 147). Yet, some re- 

 covery undoubtedly takes place as time goes on, for when one 

 set of plants both with and without surface growth is turned 

 out and examined before growth is complete, whilst another set 

 is left till growth is complete, the latter series will lead to a lower 

 estimate of the effect of the surface crop than the former. Out 

 of ten series of experiments of this sort (XVII), only one and 

 that possibly erroneous through accident gave different results ; 



