156 Discussioj'i of Coupling [ch. 



which contain erect and hooded, coloured and white, long 

 and round pollens, it was found that in 7% the distribution 

 followed the system derived from the gametic series 

 7:1:1:7, but among the F^ families derived from them 

 several were found to exhibit coupling between the blue 

 factor and long pollen according to the system 15:1:1:15. 

 For instance, to take the group of plants in which this was 

 most evident, a certain /% plant with purple flower and 

 erect standard together with eight similar plants (its offspring, 

 in F^ gave collectively 



Purple Red 



long round long round 



expectation on 583 26 24 170 



15 : I : I : 15 basis 578 24 24 177 



Here it is practically certain that the 15 : i formula 

 correctly expresses the gametic distribution. Nevertheless, 

 though some of the offspring of F^ gave such definite 

 indications of the 15 : i system others no less definitely 

 followed the 7 : 1 plan, and others again gave results so 

 uncertain that it was impossible to assign them to either 

 group with any confidence. 



The facts are thus exceedingly complex, and all that can 

 be stated is that coupling between the blue factor and the 

 long pollen does certainly exist in certain families derived 

 from a cross which involved the factors for erect and hooded 

 standard ; but that inasmuch as the F^ distribution followed 

 the 7 : I plan^, the heterozygosis between erectness and 

 hood cannot be the direct cause of this change in the 

 coupling. Pending further analysis the distinction which 

 decides whether the coupling shall follow the 7 : i system 

 or the 15 : I system must be regarded as quite unknown, 

 for examination of the various families has not revealed any 

 consistent difference between them (see 22, pp. 10—13). 



Since the B factor is alternative to E, the erect standard, 

 in the gametic composition, it follows that the zygotic com- 



