MEASUREMENT OF VARIABLE PROPERTIES 139 



sorts exist, but as often as a germ cell D meets an R, the 

 property R is latent ; and in a similar way r is latent when d 

 meets r. Therefore the 9 sorts are reduced to 4 — viz. 



The seeds DH'^ are Dd (round, yellow) 



(Compare (4) 



2DRd^ 

 2DHr 

 4DdRr 

 RH^ 

 2drR^ 



2DRr^ 



j^2y2 



120. 



Dd ( 

 Dd ( 



Dd ( „ „ ) 



Rd (wrinkled, yellow) 

 Rd ( „ „ ) 



Dr (round, green) 

 Dr( „ , ) 



Rr (wrinkled, green) 



The numerical value of each term is 

 the frequency of the corresponding sort of seeds — viz. 



Round, yellow (Dd) : ^^^^^+-^ = 1^ 



Wrinkled, yellow (i?^) : ^= A 



Round, green {Dr) : ^^= fV 



Wrinkled, green (Rr) : yV = x© 



or 9 : 3 : 3 : 1 (relative frequencies). 



In MENDEL'S experiment (556 seeds yielded by 15 Fi 

 plants) the figures were ^ : 



Observed Calculated * 



Rr (wrinkled, green) 

 Dr (round, green) 

 Rd (wrinkled, yellow) 

 Dd (round, yellow) 



Considering the four above groups with reference to their 

 hereditary possibilities, we discern : — 



Among the plants raised from round, yellow seeds : 



(a) Plants raised from seeds DH^, which are expected to 

 yield i sort of seeds (round, yellow) ; this is expressed by the 

 coefficient of the term, which is i. 



(/3) Plants raised from 2DRd'^ ; coefficient 2 ; 2 sorts of 

 seeds (round, yellow; wrinkled, yellow). 



(7) Plants raised from 2DHr ; coefficient 2 ; 2 sorts of 

 seeds (round, yellow; round, green). 



(5) Plants raised from ^DRdr ; coefficient 4 ; 4 sorts of seeds 

 (round, yellow ; round, green ; wrinkled, yellow ; wrinkled, 

 green). 



1 MENDEL, in BATESON, loc. cit., p. 333. Compare the method of cal- 

 culation adopted by Mendel and the method ifollowed here. 



^ These experimental results are very satisfactory. Compare the figures 

 obtained by tossing 500 times one coin, p. 119- 



