54 INHERITANCE IN GUINEA-PIGS. 



exactly as in the crosses between races of widely different and heritably 

 different sizes, but without indication in either case that the size inheri- 

 tance is other than a simple and permanent blend. 



THEORETICAL EXPLANATIONS OF SIZE INHERITANCE AND OF BLENDING 

 INHERITANCE IN GENERAL. 



We conclude therefore that, so far as present knowledge goes, the 

 statement made in 1909 that size inheritance is blending and does not 

 mendelize still holds. This does not preclude the possibility that in 

 special cases mendelizing factors may exist which affect size. For 

 example, in man brachydactyly is due to such a factor, a simple 

 Mendelian dominant, as was first shown by Farabee (1905), and has 

 been confirmed by Drinkwater in the case of three separate English 

 families. This character involves a shortening of the skeleton generally, 

 but of the digits in particular. It is transmitted only through affected 

 individuals, the normal offspring of affected individuals producing only 

 normals. Professor James Wilson has stated that the Dexter-Kerry 

 cattle of Ireland differ from ordinary Kerry cattle by a similar men- 

 delizing factor. If one were to restrict his study of size inheritance 

 to cases such as these, he would reach the conclusion that size inheri- 

 tance in general is Mendelian, a wholly mistaken idea. (See Castle, 

 1914.) Such cases among animals are distinctly rare. Among culti- 

 vated plants they seem to be somewhat commoner, so that many of the 

 inherited size differences studied by botanists involve such factors. 

 One of the commonest of these is involved in the difference between 

 normal (tall) and dwarf habit of growth, a case demonstrated by 

 Mendel for peas in his original experiments ; but it is more than doubtful 

 whether Mendelian factors produce the differences in height observed 

 among different races of tall or of dwarf peas respectively. The same 

 is true concerning differences in size or shape of seeds and fruits, as 

 described by Emerson and Gross. It seems almost certain that 

 Mendelian factors are involved in many of the cases studied, but 

 associated with other factors not Mendelian, possibly merely physio- 

 logical, which render the results extremely complex and the variation 

 seemingly continuous in character. To have shown that size inheri- 

 tance is occasionally affected by Mendelian factors is not by any means 

 to have demonstrated that all size inheritance is due to Mendelian 

 factors. The physiological increase of size due to the crossing of unre- 

 lated races is a fact of far greater economic importance to the animal 

 breeder than the existence of any Mendelian factor affecting size that 

 has thus far been demonstrated. 



The question may be raised, how are we to account for the increased 

 variability of F 2 as compared with F 1; if this is not due to segregation 

 and recombination of multiple factors, as assumed under the Nilsson- 

 Ehle principle. (1) This would be sufficiently accounted for in the 



