1/2 MAINE AGRICULTURAL EXPERIMENT STATION. I914. 



able and, therefore, within the limits of the experiment, less 

 affected by differences in soil and environment." 



Table 7 gives the average number of bushels per acre for the 

 four plots of each variety in 191 3. The probable error of this 

 mean is also tabled. The last two columns of the table give 

 respectively the standard deviation and the coefficient of varia- 

 tion for the yield of these plots together with their probable 

 errors. The varieties have been arranged in the table in the 

 order of the size of their coefficients of variation. The standard 

 deviation in these cases run nearly parallel with the coefficients 

 of variation so that in most instances the varieties are also in 

 the order of their standard deviations. 



"In discussing the significance of such variation constants it is always 

 necessary to take into account the statistical probable error to which 

 they are subject. This probable error in no way takes account of any 

 experimental errors. It simply tells us the amount of confidence, on the 

 basis of chance, that can be put in the given constant. Thus we know 

 that if we repeat an experiment a number of times under the same con- 

 ditions we feel certain that the average of a'll the results gives a more 

 nearly true value than any one of the results alone. The greater the 

 number of the separate trials, the more confidence one can put in the 

 average results. The probable error tells how much confidence we can 

 put in the given result. It is of such size that the chances are equal 

 that the true value lies within the limits obtained by adding and sub- 

 stracting the probable error from the calculated value. The probable 

 error is always preceded by the plus or minus (±) sign. 



The method of calculating the probable error of various constants is 

 given in the current treatises on biometrical and statistical methods. A 

 brief discussion of these is given in Part I of Bulletin no, Bureau of 

 Animal Industry, U. S. Dept. of Agr., pp. 22-24, iQio. 



