106 A MANUAL OF MENDELISM 



hens could be tested, but, as Dr. Pearl puts forward 

 the same theory, minus the suggestion, we may put the 

 suggestion aside, not necessarily with regret, since it 

 may have helped us to understand a problem so difficult, 

 and proceed to the proofs. 



The proof is parallel to that adopted for the colours 

 of cattle and horses. If, from the reproductive point 

 of view, there are nine different kinds of cocks and 

 six of hens, then, in flocks containing all kinds of hens, 

 each of the nine different kinds of cocks should leave 

 different assemblages of pullets, as indicated in the 

 table on page 103. 



The table on pages 104 and 105, taken from Dr. 

 Pearl's statistics other than those relating to second 

 and later crosses, shows, by their daughters, that eight 

 of the nine different kinds of cocks have been identified. ' 

 Figures which are not in accordance with expectation 

 are enclosed within brackets. It will be noticed that 

 several times the figure | is set down. This is to in- 

 dicate hens which laid exactly 30 winter eggs and 

 were recorded in two grades by | being placed in 

 each. 



With such close agreement between expectation and 

 actual results — the disturbances to which vital activity 

 are liable being allowed for — there can be no doubt 

 that Dr. Pearl's hypothesis is established. Another 

 factor, therefore, is found coupled with that for female- 

 ness, and, at the same time, two factors are found 

 whose effects are similar when operating separately but 

 cumulative when operating together. Thus the possible 

 constitutions in a population of hens in which all 

 the factors which have been discussed are operating 

 may be represented thus : B = barred, b = unbarred ; 



