PHENOMENA OF INHERITANCE 439 



length remained intermediate between that of the parents. He found 

 the same thing true of length and breadth of the skull (Fig. 65) and 

 of the size of other portions of the skeleton, and he concluded that such 

 quantitative characters are not inherited in Mendelian fashion. 



Quite recently MacDowelL working on the inheritance of size in 

 rabbits, concludes that this, as well as other quantitative differences be- 

 tween parents which appear to blend in the offspring, such as Castle's 

 case of ear length in rabbits, is due not to a single factor, as in the 

 case of Mendel's tall and dwarf peas, but to several factors. Conse- 

 quently, in the formation of the germ cells there is not a clear segrega- 

 tion of all the factors for tallness, or large size or long ears, in half the 

 germ cells, and their total absence in the other half of those cells, but 

 some of these factors go into certain cells and others into others, as 

 in the case of dihybrids, trihybrids or polyhybrids. As a result off- 

 spring appear more or less intermediate in size between their parents. 



Thus it is possible to explain even " blending " inheritance as due 

 not to the real fusion or blending of inheritance factors, but to varying 

 combinations of numerous or multiple factors, according to the Men- 

 delian rules. The Mendelian principle of segregation has been found 

 to be of such general occurrence that there is a strong inclination among 

 Mendelians of the stricter sort to make it universal, and to explain all 

 cases of blending inheritance as due to incomplete dominance and to 

 multiple factors. Whether or not such attempts may prove completely 

 successful it is still too soon to say. 



III. Mexdeliax Inheritance in Max' 



The study of human inheritance must always be less satisfactory 

 and conclusions less secure than in the study of lower animals for the 

 following reasons: In the first place there are no "pure lines," but the 

 most complicated intermixture of different lines. In the second place 

 experiments are out of the question and one must rely upon observa- 

 tion and statistics. There have been less than 60 generations of men 

 since the beginning of the Christian era, whereas Jennings gets as 

 many generations of Paramecium within two months and Morgan al- 

 most as many generations of Drosophila within two years. Finally 

 the number of offspring are so few in man that it is difficult to deter- 

 mine what all the hereditary possibilities of a family may he. Bearing 

 in mind these serious handicaps to an exact study of inheritance, it is 

 not surprising that the method of inheritance of many human charac- 

 ters is still uncertain. 



Davenport and Plate have catalogued more than sixty human traits 

 which seem to be inherited in Mendelian fashion. About fifty of these 

 represent pathological or teratological conditions, while only a relatively 

 small number are normal characters. This does not signify that the 



