:?3G Relatione hetween Growth and Envlromnent 



When these figures are compared with those for the "changed" 

 series given on page 226, it is seen that with the young plants little differ- 

 ence occurs, showing that in early life the amount of food supplied in 

 both sets was in excess of requirements and did not therefore act as a 

 limiting factor. With older plants, however, when the food solutions 

 were not renewed the correlation of rate of increase with mean maximum 

 temperature was very much lowered and that with bright sunshine was 

 also reduced to a significant extent. This indicates that when scarcity 

 of food is acting as a limiting factor the plants are unable to take full 

 advantage of the available bright sunshine, while at the same time the 

 beneficial influence of high maximum temperature is reduced to an even 

 greater degree. 



Another interesting point can be elucidated from the comparison 

 of seven pairs of changed and unchanged series, each pair being grown 

 under identical conditions for the same length of time. The total 

 variance^ at any given age, for older plants (in the second period of 

 growth), may be roughly analysed into 



37 per cent, fortuitous variation (due to individual difference of 

 seeds etc.). 



38 per cent, known environmental factors. 



25 per cent, unknown environmental causes (irregularities of 

 temperature and light not affecting the means, humidity etc.). 



The percentage of variance (38 per cent.) that can be attributed to 

 the known environmental causes of maximum and minimum temperatures 

 and total sunshine is remarkably high and indicates that these are the 

 most potent of the factors acting upon growth. The causes of fortuitous 

 variation are less clear, and the percentage (37 per cent.) seems rather 

 high considering that it is a mean figure, as under the greenhouse con- 

 ditions the extremes of individual variation in a number of plants 

 generally reach about the same figure, and the mean variance in the 

 rate of increase might be expected to be less. The extent of individual 

 variation in weight, however, regarded as a percentage difference, must 

 not be confused with the variance of the efficiency index which is to be 

 ascribed to various groups of causes. F\nther work on this point will 

 be necessary, for if it could be more fully exj)lained it would give more 

 reliable information than is at present available as to the influence of 

 the individuality of plants upon the validity of experimental results. 



' Tlic mean H(iuare deviation of the rate of increase at given age is used as a measure 

 of the variance, and the percentage figures are caleuhited from tills. 



