418 



THE AMERICAN NATURALIST [Vol. LV 



tribution of the young obtained when Fj black-eyed- 

 whites from the above cross were mated with pink-eyed 

 piebalds, the double recessives. The cross-over classes 

 number 134, or 46.68 ± 1.99 per cent, of the total 281 ani- 

 mals raised. The deviation from the distribution ex- 

 pected is within the probable error and indicates inde- 

 pendent segregation. 



The reciprocal cross w^as also made in w^hich the two 

 genes entered together in the same parent. The Fj 

 gametes should again consist of equal numbers of two 

 classes, WP and wp (the cross-overs), and wP and Wp 

 (the non-cross-overs), if independent segregation occurs. 

 If linkage is shown, the latter class should be larger. 

 The actual distribution is shown in Table III opposite 

 "new data" of the coupling series. The agreement with 

 expectation is poor due to small numbers (77). But since 

 the cross-overs exceed the non-cross-overs, it is clear that 

 no linkage exists. The cross-over per cent, is large 

 (59.74 ±3.77), but the deviation is still less than three 

 times the probable error. By combining both coupling 

 and repulsion results recorded in the new^ data, a total 

 of 358 animals is obtained, of which 180 are cross-overs. 

 This gives a cross-over percentage of 50.27 ± 1.78. The 

 deviation from the normal distribution is 0.27 per cent., 

 a figure well within the probable error. 



These crosses, as well as Dunn's, show^ consistently 

 the independence of tlic black-eyed- white and pink-eye 

 genes. Further evidence may be had by adding together 

 the cross-over and non-cross-over classes of the two sets 

 of data as shown in Table III. The cross-overs number 

 330 and constitute 51.48 ±: 1.32 per cent, of the total 641 

 animals raised. The deviation from the expected dis- 

 tribution is 1.48 and not significant. The number of 

 young raised does not equal that of other crosses, for 

 the consistency with which these data, in agreement with 

 DetlefseiTs ol)>crvations, show the indepeihli lu-r nf llie 

 two Irenes, warranted the discontinuance ..f fnrtlier 

 breeding. 



