R. C. PUNNETT AND P. G. BaILEY 



211 



dnced from crosses with clean-shanked breeds such as the White 

 Leghorn and Hamburgh all shewed leg-feathering, though in less degree 

 than the Cochin parent. At the time of Hurst's experiments the idea 

 of multiple factors had not yet been put forward. It was not at that 

 time fully realised how important it was to keep entirely separate the 

 record of each bird used as a parent. Many similarly bred hens were 

 run with a given cock on the assumption that they were of like con- 

 stitution though, as we know to-day, this was not necessarily the case. 



Nevertheless Hurst's results can to some extent be reconciled with 

 those of Davenport. In Exp. 6 (p. 135) a number of F^ birds were 

 raised from (o) 2 Hamburgh $ $ x Buff Cochin (/ and (b) from Buff 

 Cochin $ X Hamburgh ^. The 107 birds reared all had feathered 

 legs, though the feathering was distinctly less than that of the Cochin. 

 In Exp. S (p. 1.38), 12 of these F, % % were mated with an F, ^, and 

 gave 11.5 chicks with feathered and 4 with clean legs. 



Assuming that all of these birds were AaBb in constitution expecta- 

 tion would be 111-5 £1. : 7-5 n.f.l. Twelve i^i % $ (presumably different 

 ones from the above since the experiment vvas also made in 1903) were 



