No. 639] IXHERITANCE OF BLUE IN POULTRY 305 



These results are in substantial accord with those of 

 Bateson and Punnett (1906, p. 20). A somewhat marked 

 departure from the theoretical expectation appears in 

 group (3) of black c^d" X blue ?? matings, the agreement 

 in the reciprocal cross (group 2) being as close as pos- 

 sible. This departure from expectation is due to the 

 progeny of a single pair of birds (J'lSeM and $2005) 

 which produced 25 blues and 6 blacks. If the latter are 

 left out of consideration the results are 88 blues and 

 84 blacks. 



How^ever, even in the case of the progeny of d'lSGM and 

 $2005 the Dev./P.E. = 4.1, which indicates a deviation 

 of doubtful significance. The results of this mating were 

 carefully considered from the standpoint of crossing- 

 over, but there is no indication of its having occurred. 



According to these results the genetic compositions of 

 the three color types of Andalusians used in these ex- 

 periments were as follows: blue-splashed = (i?6) (i?e), 

 blue = {Re){rE), and black = {yE){rE). There was no 

 evidence of crossing-over between R and E having 

 occurred. 



VI. Data from Crosses of Andalusians with 

 Certain Recessive White Breeds 



h\ the ])n"vi()us ])api'r (liHS./. ]>. IOC) the writer re- 

 ported a small nmulu'r of data on a cros> lu-twecn a wliite 

 WyaiKlottc o and a bhie-splaslu'd Aiidahisiau These 

 have been eolisiderably increased in amount and the rc- 

 eii)roeal eross made. Further, both blue and bhick Anda- 

 lusians ]in\ e been crossed reciprocally with white AVyan- 

 d()tte> and all three Andalusian color tyix'S crossed 

 reei])rorally with white Plymonth l?oeks. The data from 

 these srwral niatin,u> ar.. s.^t iorth in Tahlr II. 



The fro>srs wrre luixdv in the twvlve possihh. wavs, 

 from eh'ven of wliieh offsprin- were s.. aired, th. onr type 

 of matin,- which failed to prodiu-., oflVprin- h-'in- the 



