314 Appendixes 



capable of occurring at any cell-division, and not merely in 

 gameto-genesis. The moment of segregation will thus 

 probably be found to vary with different types and with 

 different factors, and for some it is likely that great 

 irregularity in this respect will be found as a common 

 occurrence. In this way can we best hope to reconcile the 

 facts that with certain types such as wheat and barley the 

 expected ratios are almost always approximately realised, 

 whereas in other types, such as Matthiola there is great 

 fluctuation in the numbers from individual pods, as 

 Miss Saunders assures me, though when large series are 

 taken the normal ratios are usually approached. 



Appendix to Chapter IX. 



The result of further experiment has led to a great 

 simplification of the views as to coupling and repulsion 

 given in the text. We have now proof that coupling and 

 repulsion are consequences which may affect the same pair 

 of factors. If AB x ab gives an F^ in which A and B are 

 coupled, then Ab x aB will give F^ in which A and B are 

 "repelled." It is moreover established in three cases that 

 the appearance of "repulsion" is in reality a consequence 

 of the fact that the gametes Ab and aB are represented in 

 numbers larger than those in which AB and ab occur. In 

 other words, if AB x ab gives F^ AaBb having gametes 



jAB : lAb : \aB : : ja6 ; 

 then Ab x aB will give F ly AaBb having gametes 



lAB : jAb : jaB : lab, 



or at least displaying some similar coupling. We have as 

 yet no sufficient data for asserting that the actual numbers 

 will be the same in the two cases, though this is perhaps 

 the simplest expectation. 



The essential phenomenon in each case is the formation 

 by the heterozygote of a greater number of germ-cells repre- 

 senting those combinations of factors which were possessed 

 by the original parents, and fewer germ-cells representing 

 the new combinations*. 



* I have only lately seen Emerson, R. A., "Genetic Correlation and 

 Spurious Allelomorphism in Maize," 2$th Ann. Rep. Nebraska Agric. 

 Exp. Sta. 1911, in which this possibility is pointed out already. 



