174 PHYSIOLOGIC GENETICS 



anomalous eye rosette and born alive, 96 were recorded as having a strong pattern 

 and 75 a weak one. There was rather frequent irregular asymmetry, including a few 

 cases in which there was a strong rosette about one eye and none about the other. This 

 indicates that developmental accident played a considerable role. 



In matings in which Mm and MM were segregating, there was no significant 

 difference in incidence of eye rosettes among grades E, D, and C contrary to expectation 

 in view of their invariable occurrence in grades AC and A, genotype Rmm. There was 

 also no relation to presence or absence of a forehead rosette in matings segregating in 

 Stst and stst, discussed later. On the other hand, segregation of R and rr made a 

 great deal of difference. Thus a group of matings that produced 16 per cent with 

 eye rosettes among 208 RM young produced only 2 per cent among the 217 rr young. 

 A group that produced 55 per cent among 88 RM young, produced only 20 per cent 

 among the 46 rr segregants. Only liveborn are considered here because of somewhat 

 greater uncertainty in classification of stillborn. Thus R greatly favors manifestation 

 of the character but is not necessary. Strong eye rosettes were present in eleven other- 

 wise completely smooth animals (rr). 



The highly sporadic occurrence at first, including such records as 1 in 15, 16, or 

 17 RM young from scattered matings between normals, of in 27 and 2 in 29 from 

 matings of strong x normal and of 3 in 15 from a mating of strong x strong, suggested 

 either low penetrance or a recombination effect of two or more genes. After all young 

 were examined carefully for weak eye rosettes at both birth and weaning (when they 

 are often more conspicuous), it became apparent that penetrance at this level could 

 be high at least in RM young. The pattern of occurrence indicates a single essential 

 gene, Re (rough eye) . The results of outcrosses indicated dominance in some degree 

 but with penetrance far from perfect in heterozygotes carrying RM and low but not 

 zero in ones carrying rr. There was an approach to 100 per cent penetrance but not 

 expressivity of RMReRe, but only moderate penetrance in rrReRe. 



Another aspect of the full rough pattern of Rmm, the forehead rosette, can also 

 appear in the absence of R. li29, 1430 All animals in the colony of this type traced 

 to three obtained from Mr. I. J. Wachtel. This character, Star (St), behaved in a 

 remarkably different way from rough eye (Re) in two respects. First, it showed no 

 intergradation with normal and behaved as a simple dominant in eight successive 

 backcrosses to lines in which it had never been present and there was no difference 

 between rrStSt and rrStst. li29 - 1430 Second, instead of showing enhancement in the 

 presence of R there was a tendency toward reciprocal inhibition. 1430 In rrSt-, the 

 forehead rosette is almost invariably (99.6 per cent) single and very flat. In about 

 19 per cent of RMMSt it is replaced by two weak rosettes. This forehead roughness is 

 combined with typical grades D or E but the proportion that are grade D is reduced. 

 In RMmSt-, this reduction of the rough pattern goes farther. Most show roughness of 

 grades D or even E combined with a forehead rosette that is doubled in 64 per cent. 

 In the 1 7 per cent called grade C the rosettes were usually weak and lateral. The 

 forehead rosette was weakened in 78 per cent. Finally in RmmSt in which an exception- 



