Growth and variation in maize. 143 



In g-eueral Figs. 13 to 15 do not show such a gi-adual change 

 from the relatively small or relatively large plants at one end of the 

 season towards the mean of the population at the other as was noted 

 in Figs. 10 to 12. Instead there is a tendency to maintain about the 

 same relative position until near the end of the season. Then there 

 is a rather sudden change in the mean quintile position. An exception 

 to the above statement is found in the plants ending in quintile I in 

 Fig. 13. In this case there has been a gradual change from a condi- 

 tion near the mean to relatively small plants at the end of the season. 

 It was pointed out above that the quintile in which a plant ends 

 is not as good a measure of its average relative size for the season 

 as is the quintile in which it started. A comparison of these two sets 

 of figures aids us in getting at some of the causes of this. Probably 

 the chief reason lies in the fact that certain plants which are relative- 

 ly large, or at least of medium size, throughout the season cease 

 growing earlier than other plants of the same size. Consequently in 

 the final measurements these plants are relatively small compared with 

 the remainder. Likewise some plants have a longer growing season 

 and, on that account, become relatively large in the last measurements, 

 although for the remainder of the season they were relatively small or 

 of medium size. 



These facts undoubtedly account for many of the peculiarities in 

 Figs. 13 to 15. 



The mean quintile position of individual plants. 



It is next proposed to carry the analysis of the growth cui-ves 

 one step farther and to consider the quintile distribution of the mea- 

 surements of each individual plant. The quintile distributions for each 

 plant in each series are shown in tables 61, 62 and 63. The plants 

 have been arranged in these tables for a special purpose, but since 

 the individual number of each plant is given, they can serve as the 

 tables of fundamental distributions. The mean quintile position^) and 

 its standard deviation have been calculated for each plant in these 

 distributions. These constants were calculated by the method of nio- 



') It should be noted that the ,mean quintile position' of an individual plant 

 is entirely different from the term ,mean quintile position' used in the preceding section 

 of this paper. As used in the remainder of this paper the term ,mean quintile posi- 

 tion' will refer to the individual plant unless expressly stated to the contrary. 



