178 E. C. MACDOWELIi 



tain strains he obtained consistent ratios which indicated di- 

 and tri-hybrid crosses (15 : 1, 63 : 1); in other words, there were 

 two or three factors involved, each of which alone or in combi- 

 nation with the others, produced the same color, and only in 

 the absence of all such factors could the recessive color be seen. 

 The application of this theory to the case of the rabbit ear 

 lengths was immediately made. East ('10), Castle ('11), Lang 

 ('11), by supposing that ear length depended upon several fac- 

 tors each of which behaved as a simple Mendelian unit, but lacked 

 dominance. This would mean that we could see the difference 

 between an ear whose length had been based upon four factors, 

 and one whose length depended upon five or six factors. It 

 would be supposed that the more simplex factors, or doses, there 

 were present, the longer the ear. If the long-eared race bore 

 four duplex factors (8 doses) for length, and the short-eared race 

 bore two similar but not allelomorphic factors (4 doses), the hy- 

 brids obtained by crossing these two races would each have six 

 factors in a simplex condition (6 doses). The relative sizes of 

 the two parents and the hybrids would be expressed by the num- 

 ber of doses in each case. The second generation would give rab- 

 bits with various numbers of doses and their ears would range 

 all the way from the length of the long-eared race to that of the 

 short-eared race. Unless the two factors in the small race were 

 allelomorphic to two of the four in the large race, there could be 

 found rabbits with ears longer than the long-eared race, and 

 others with ears shorter than the short-eared race, for in the sec- 

 ond generation would be formed combinations involving as many 

 as 12 doses and others with none. There would, however, be a 

 much greater number of rabbits with intermediate than with 

 extreme ear-lengths. Just as in a mono-hybrid cross without 

 dominance there are two heterozygous individuals to one of each 

 of the pure types, so in a multo-hybrid cross, where each factor, 

 independent of the others, is giving a 1 : 2 : 1 ratio, the individ- 

 uals with all the factors in a simplex or heterozygous condition 

 would be most numerous. The higher and lower grades would 

 appear in decreasing frequency as the extremes were approached 

 (Tammes '11). Now if only a few animals were raised in F2 



