160 



Pearl and Surface. 



Table 

 Showing the mean and standard deviation of the total quintile distri- 



positions in the 



In the third from the last row of this table the constants for the 

 thi-ee series are averaged. There is no essential difference in respect 

 of these constants between the three series so that the average of the 

 three will give a better value than to take each series separately. 



From this table it is noted that in general the average means 

 are not widely different from the theoretical means. Tliis is especially 

 true of the middle classes where the differences between the average 

 and theoretical means are but little greater than the probalMe errors. 

 In the end classes the differences are several times the prol)able errors 

 and hence are undoubtedly significant. It is further noted that in the 

 case of small plants the observed means are smaller than the theore- 

 tical while with large plants the reverse is true. 



In regard to the standard deviations the differences are much 

 more marked. In every case these differences are from ten to fifteen 

 times their probable error. Further in every case the observed stan- 

 dard deviations are lower than the theoretical. Thus the measure- 

 ments of the corn plants are much less scattered about their 

 means than the most probable distributions on the theory of 

 chance. 



') Tlie probable errors of these averages are calculated by taking n = the sum 

 of the variates in the three series. 



