194 



SCIENCE 



[N. S. Vol. XLII. No. 1075 



pigmented (gray or black) rats, wliich when 

 crossed with each other produced an F^ gen- 

 eration consisting of approximately 9 gray or 

 black rats to 3 yellows, and 4 albinos, a typical 

 dihybrid result. 



But the most interesting aspect of the two 

 yeUow variations is their apparent "negative 

 coupling " or mutual " repulsion," a phenome- 

 non first discovered by Bateson and Punnett in 

 plants, but since observed in insects by Mor- 

 gan and Tanaka. This repulsion, in the case 

 of the two yellow variations of rats, is incom- 

 plete, as is true in most recorded cases of re- 

 pulsion, and it indicates, if we adopt Morgan's 

 manner of explaining it, a location near to- 

 gether in the germ-plasm of the respective de- 

 terminers or " genes " for red-eyed and for 

 pink-eyed yellow. 



The evidence for this partial repulsion is as 

 follows: When the Fj gray or black rats, ob- 

 tained by intercrossing the yellow varieties, 

 were themselves intercrossed, they produced 

 an Fj generation which contained 102 gray or 

 black young, 55 red-eyed yellows, and 43 pink- 

 eyed yeUows. 



To ascertain the gametic composition of the 

 yellow Fj animals, (1) the extracted red-eyed 

 yellows were mated with yellows of the pure 

 pink-eyed race, and (2) the extracted pink- 

 eyed yellows were mated with pure red-eyed 

 yellows. Twenty-eight test matings of the 

 first sort have been made and twenty-seven of 

 the second sort, with the following results. 



In 20 matings, extracted red-eyed yellows 

 mated with pure pink-eyed yellows have pro- 

 duced only black-eyed young (grays or blacks), 

 while in the remaining 8 matings both black- 

 eyed young (grays or blacks) and pink-eyed 

 young have been produced. The red-eyed pa- 

 rent, in the former type of mating, must have 

 lacked altogether the gene for pink-eye, while 

 in the latter type of mating it must have been 

 heterozygous for pink-eye. If we designate 

 the (recessive) gene for red-eye by r and the 

 (likewise recessive) g&ne for pink-eye by p, 

 then the Fj red-eyed yeUows tested must evi- 

 dently have been of the two types rr (20 eases) 

 and rrp (8 cases), respectively. The former 

 type, when mated with pure pink-eyed animals 



{pp), would produce only double heterozygotes 

 {pr), whereas the latter would produce either 

 double heterozygotes (pr) or homozygous pinks 

 heterozygous for red-eye (ppr). On the theory 

 of probability, if red-eye and pink-eye are pro- 

 duced by wholly independent genes, we should 

 expect animals of the former type to be only 

 haK as common as animals of the latter type 

 in a population of extracted (Fj) red-eyed yel- 

 lows, iut observation shows them in this case 

 to he more than twice as common! Hence 

 there is a strong presumption that red-eye and 

 pink-eye depend upon genes not wholly uncon- 

 nected with each other. This presumption is 

 strengthened by the results obtained by test- 

 ing the extracted F„ pink-eyed yellows by 

 mating them with pure red-eyed yellows. 

 Twenty-seven such animals were tested, of 

 which 19 produced only black-eyed young 

 (grays or blacks), while 6 produced both black- 

 eyed (gray or black coated) and red-eyed (yel- 

 low) young, and 2 others produced only red- 

 eyed (yellow) young. 



All the pink-eyed F, animals tested must 

 have been homozygous for pink-eye (otherwise 

 they would not have shown pink-eye), but as 

 regards the possession of red-eye it is evident 

 that conceivably they might (1) lack it alto- 

 gether, (2) be heterozygous for it, or (3) might 

 be homozygous for it, conditions which would 

 be expressed by the formulse pp, ppr and pprr, 

 respectively. Animals of the first sort (pp), 

 if mated with pure red-eyed animals, should 

 produce only black-eyed young (grays or 

 blacks, the observed result in 19 cases) ; ani- 

 mals of the second sort (ppr) should produce 

 some young black-eyed and others red-eyed 

 (the observed result in 6 cases) ; animals of 

 the third sort (pprr) should produce only red- 

 eyed yellow young, an expectation realized in 

 2 cases. The chance expectations for the oc- 

 currence of these three sorts 'of results are as 

 1:2:1; the observed occurrences are widely 

 different, viz., 20 : 6 : 2. 



If the repulsion between the two yellow vari- 

 ations were complete, no Fa individuals of 

 classes 2 and 3 (ppr and pprr) would be formed, 

 but all F: pink-eyed yellows would be of class 1 

 (pp). The fact that classes 2 and 3 are 



