GAMETIC COUPLING IN YELLOW RATS. 



177 



decide between these possibilities. Further, it was clear that every 

 pink-eyed F 2 yellow must be homozygous for that character, which is 

 also recessive, but might be either homozygous or heterozygous for 

 red-eye without affecting its appearance, or might even lack the gene 

 for red-eye altogether. A cross with pure red-eyed animals would 

 suffice to show in each case which possibility was realized. In accord- 

 ance with this reasoning the proposed tests have been made in the case 

 of 45 red-eyed and 40 pink-eyed F 2 yellows. 



Of the 45 red-eyed yellows tested, 32 have given exclusively black- 

 eyed young (blacks or grays), no test being considered adequate which 

 did not produce 4 or more young; but 13 of the tested animals gave a 

 mixture of black-eyed and of red-eyed young in approximately equal 

 numbers. The former group, numbering 32, evidently lacked the gene 

 for pink-eye, since they always produced atavists in crosses with pink- 

 eyed yellows; the latter group, numbering 13, were evidently hetero- 

 zygous for pink-eye, since only part of their young were atavistic. 



Of the 40 pink-eyed F 2 yellows which were tested, 27 produced only 

 black-eyed young; these evidently lacked the gene for red-eye. Ten 

 others produced both black-eyed and red-eyed young, being evidently 

 heterozygous for red-eye. Three have produced only red-eyed young, 

 which shows them to be homozygous for red-eye as well as for pink-eye. 

 Hence they are the double recessives, expected to be one-sixteenth of 

 all F 2 rats if no linkage occurs, but less numerous if linkage occurs. 



We are now in a position to estimate the strength of the linkage 

 shown. If we designate by r the recessive gene for red-eye, and by p 

 the recessive gene for pink-eye, then in the current Mendelian terminol- 

 ogy the following F 2 classes are to be expected in the frequencies shown, 

 if no linkage occurs : 



For the present we may pass by the black-eyed classes, since none of 

 these were individually tested. The individual tests already described 

 have shown the existence of the expected two classes of red-eyed arid 

 three classes of pink-eyed young, but in proportions very different 

 from those given in the table. Among the red-eyed, instead of the 

 expected 1 rrPP : 2 rrPp, we observe 32 : 13. Among the pink-eyed, 

 where we expect 1 RRpp : 2 Rrpp: 1 rrpp, we observe 27 : 10 : 3. These 

 are very different frequencies from those expected, and they strongly 

 suggest linkage. How strong is the linkage? We may estimate it 



