120 Pearl and Surface. 



butious for the successiTe measurements, first of the plants which start 

 in given quintiles and second of the phmts which end in given quintiles. 

 These distributions are given in tables 31 — 60. In these tables there 

 are also given the mean quintile position, and the standard deviation 

 of the quintile distribution, of the plants for each measurement. It will 

 be more convenient to discuss these constants after certain other 

 features of these distributions have l)een considered. 



The quintile deviations of plants starting in a given qnintile. 



In tables 31 — 45 there is given in each case the total number of 

 observations which fall in a given quintile for the whole season, ex- 

 clusive of the measurements at the beginning of the season. In Table 31, 

 for example, there were 11 plants which were in quintile I of the 

 frequency curve resulting from the first measurement, June 12. On 

 June 15 seven of these plants were again in quintile I of that distri- 

 bution, three were in quintile II and one was in quintile EU. Now if 

 we exclude the measurement on June 12, because, by hj-pothesis, all 

 these plants then fell in quintile I, there are thirteen distributions 

 remaining, each containing eleven observations, or in all 143 observations. 

 From the sums at the foot of tliis table it is seen that of these 143 

 observations, 69 fell in quintile I, 41 in quintile II, 20 in quintile HI, 

 13 in quintile IV and none in quintile V. Or, transferring to percentage 

 figures, 48'25 percent of the observations made on these plants fell in 

 quintile I, 28'67 percent in quintile 11, and so on. Stating it in another 

 way, it may be said that the eleven plants starting in quintile I spent 

 half their time (48'25 percent) in that same quintile, and only one- 

 fourth their time (28'67 percent) in (luiutile II, and still less in any one 

 of the remaining quintiles. 



There is, then, a marked tendency for the plants which were 

 relatively small at the beginning of the season to have remained, on 

 the average, relatively small throughout most of the season. In order 

 to bring these facts for the several tables into a more condensed form, 

 as well as to show certain other relations, we have collected these 

 figures for plants starting in each quintile, and for each series in 

 tables 7, 8 and 9. In these tables there are given both the actual and 

 percentage frequencies of the observations falling in the several quin- 

 tiles, for plants starting in the given quintile. Thus the first row of 

 table 7 shows for series A, the number and the percent of observations 

 falling in the different quintiles for the plants which start in quintile I. 



