1908.] Theory of Ancestral Contributions in Heredity. 69 



The actual deviation of the percentage of wrinkled s from expectation is 

 + - 97 ; the Probable Error of this percentage is + 0-498 : so that the actual 

 deviation is just less than twice as great as the Probable Error, and is 

 certainly not significant. Moreover, it is in the opposite direction to that 

 which would afford any support to the theory of Ancestral Contributions. 





Number 



Nature of parent 



Number 



Number 





of 



with regard to 



of 



of 





family. 



cotyledon colour. 



round. 



wrinkled. 





124. 1 



DR 



249 



96 





124. 2 



DR 



219 



64 





124. 3 



DE 



360 



125 





165. 1 



DR 









165.2 



DD 









165. 3 



DD 









165. 4 



DR 



393 



142 





218. 1 



DR 



162 



61 





218. 2 



DR 



408 



147 





218. 3 



DR 



219 



75 





218.4 



DD 









218.5 



DD 







TJnclassifiable families ... 



158. 1 



DD 









158.2 



DR 









158. 3 



DR 









158. 4 



DR 



297 



103 





158. 5 



DD 









158. 6 



DR 



298 



101 





175. 1 



DR 









175.2 



DD 









175.3 



DR 









175.4 



DD 













2605 



914 



Percentage of greens 25'97 



Probable Error of percentage ± 0"498 



The number of families (falling into the categories classifiable by yellow 

 and green grandparents) in which segregation occurred was 8 ; the number 

 of families from the same parents in which segregation did not occur was 4. 

 There should, on the Mendelian hypothesis, be equal numbers of the two 

 (DE and DD give 50 per cent. DE, and 50 per cent. DD). But I do not 

 think it desirable to base any conclusion on numbers so small as these. 



Control Crosses. 



The evidence on which the statement that the greens occur in the pro- 

 portion of 25 per cent, in the F 2 generation from a cross between a pure 

 yellow and a pure green pea is based, has been collected by Mr. Lock in his 

 most useful paper on " The Present State of Knowledge of Heredity in 



