78 Maine: agricui,turai, e;xperimi;nt station. 1909. 





^s^ 



^ 























\ 



/ 



\ V 























li^\ 



/ 





\ 























/ 



A 



/ 





\ 























1 





























// 



Y 



























/^ 

























/ 



V 



// 

























y> 



y 



\i 



// 











"\l 











^ 

















\ 



s 































/ 



><- 



">/i 





X 



f 













^ 



V > 1 



/ i 



r' N, 



/^ ^ i 



1 • 

















\ 



\ 



/y 







^■4-" 

















\ 



\ 



\ 



9 

















^ 



^ 



2.5 8.5 14.5 20.5 26.5 32.5 38.5 "^4.5 50.5 56.5 62.5 



£"<r<r rF'ODUCTlON 

 Fig. 13. Diagram comparing the spring (March i-June i) 

 ■%gg production of "registered" and "unregistered" pullets. 

 Lines as in Fig. 12. 



ally somewhat more variable (i. e., less true to type) relatively 

 than the group whose ancestry had not been so closely selected. 

 The exception to this rule occurs at the point where the excep- 

 tion in the rule regarding the means occurred, namely, in respect 

 to the March to June ^gg production. The "registered" birds 

 were less variable than the "unregistered" birds kept in the 100 

 bird flock. This exception, however, only emphasizes the gen- 

 eral rule which is to be observed in the other cases. In regard 

 to absolute variability, as measured by the standard deviation, 

 the "registered" birds show a very slightly smaller "scatter" than 

 do the "unregistered" birds. This is to be expected because of 

 the smaller average production of the "registered" birds. The 

 coefficient of variation is clearly the significant measure of 

 variability in the study of ^gg production, when groups of dif- 

 ferent average productivity are to be compared."* 



3. It will be noted from the values of the probable errors 

 given in the table that some of the individual differences, par- 

 ticularly those between the coefficients of variation can not, 

 when taken singly, be considered significant in comparison with 

 the probable errors. From the practical standpoint this is more 

 than offset by the fact that the differences, whether of means 

 or of relative variabilities are, with the two exceptions noted, 

 all in the same direction. 



*C£. Pearl, R. and Surface, F. M. loc. cit. 



