BEHAVIOUR OF PLANTS IN MASSES 317 



was reduced by means of series of experiments wherein different 

 intervals were allowed so that the first sown plants in both 

 sets of pots were of the same degree of vigour when the second 

 sowings took place. The results of three series of experiments 

 showed that the younger plants made only half the growth 

 (57, 48 and 52 per cent.) when subject to the interference of the 

 roots of the older plants in the undivided pots, that they did when 

 not subject to such interference in the divided pots. This 

 might be explained, either by the older and stronger plants 

 monopolising such nourishment as there was in the soil an 

 explanation which presents certain difficulties or by the younger 

 plants suffering from the toxin produced by the older plants, 

 which, of course, they would not do when each plant was 

 grown in a separate compartment. To settle between these 

 alternatives, the experiment was modified : undivided pots were 

 alone used, and the intervals between the sowings of the first 

 three and second three plants was varied from o to 20 days, 

 another set being added in which there were no second-sown 

 plants at all. The mean results from two series of experiments 

 with mustard, done in different seasons, were 



Interval Relative weight Weight of 



between the per plant. total , 



sowings. First plants. Second plants, crop. 



o 100 -f 100 = 200 



4 days . . . 81 1 + 74 = 155 



8 . 107 + 72 = 179 



12 141 + 47 = 188 



16 147 + 38 = 185 



20 174 + 18 = 192 



No second sowing . . 198 + o = 198 



Now, it has been shown that when different numbers of 

 plants of the same degree of vigour are grown in a limited and 

 insufficient quantity of soil, the amount of growth obtained is 

 the same whatever be the number of the plants ; the soil is 

 exhausted, and the nutrient in it is divided equally between the 

 plants; if there are twice as many plants present in one case 

 as in another, each plant will receive only half the supplies, 

 and form half the growth, as in the other case (p. 315). This 

 was evidently the condition obtaining in these experiments, 

 for the first and last entries the total weight of growth 



1 This low value is attributable to the results in one series only; the 

 other series gave 98 -j- 64 = 162. 



