66 Mr. Darbishire. Experimental Estimation of the [Dec. 7, 



11 groups, 5 exhibit a deviation in a — direction, and 6 in a + direction. 

 Also that 7 of the deviations (478, 479, 481, 482, 493, 483, and 484) fall 

 within the Probable Error and 4 outside, in only one of which is the deviation 

 greater than thrice the Probable Error — viz., in 480 in which the actual 

 deviation is a little over three and a-half times the Probable Error. The 

 application of this test, therefore, justifies us in concluding that the deviations 

 from 25 per cent, which occur in each of these 11 groups are attributable to 

 chance. 





Catalogue 

 number* 

 of grand- 

 parent. 



Number 



of 

 yellows. 



Number 



of 

 greens. 



Per- 

 centage 



of 

 greens. 



Actual 

 deviation 

 of per- 

 centage. 



Probable 

 Error of 



per- 

 centage. 



Classification by yellow 

















477 



14,081 



4,532 



24 35 



-0-65 



±0-212 



478 



7,172 



2,405 



25 -11 



+ 0-11 



±0 -299 





479 



20,761 



6,951 



25 -08 



+ 0-08 



±0 -176 





480 



3,978 



1,224 



23 53 



-1 -47 



±0 -397 





481 



2,777 



942 



25 -33 



+ 0-33 



±0-481 





482 



11,140 



3,715 



25 -01 



+ 0-01 



±0-240 





492 



14,157 



4,658 



24-76 



-0-24 



±0 -212 





493 



1,408 



486 



25 -66 



+ 0-66 



±0 -677 



Totals 





75,474 



24,913 



24 -82 



-0-18 



±0-092 



Classification by green 

















483 



38,565 



12,788 



24-9 



-o-i 



±0 -129 



484 



17,150 



5,719 



25 -01 



+ 0-01 



±0 193 





485 



19,759 



6,406 



24-48 



+ 0-52 



±0-179 



Totals 





75,474 



24,913 









* For signification of these, see pp. 62 and 63. 



The totals have been added up and written under the grouping by yellow 

 grandparents and under the grouping by green grandparents. But tbe per- 

 centage of greens, the actual deviation and Probable Error of the percentage 

 calculated from these totals, have only been written once, i.e., under the 

 grouping by yellow grandparents. 



A Special Case. 



In the J?2 families raised from five of the crosses there occurred 

 wrinkled as well as round seeds. All these occurred in crosses made with 

 the extracted green No. 484, which is the only one of the three races 

 of extracted greens used, from which, on the Mendelian theory, wrinkled 



