80 THE MENDEL JOURNAL 



the persistency with which a recessive character 

 breeds true when once it has been segregated out from 

 association with a dominant character. It is true 

 that, with the exception of one remarkable case, 

 which we shall consider later, we have no very long 

 continuous pedigree of an extracted recessive charac- 

 ter. But we have a number of short pedigrees, 

 and when all their indications point in one direction, 

 and almost invariably show that an extracted reces- 

 sive character breeds true when the individual 

 manifesting it is mated with another also showing 

 the recessive character, we are justified in accepting 

 this as clear evidence of segregation. And the 

 justification is not lessened because a few exceptions 

 occur. These may receive an explanation when we 

 know what it is that determines which of two alter- 

 native characters shall be the dominant one, when 

 we have tested the validity of past records by a 

 more searching examination of future cases, and 

 when pathologists and physiological chemists can 

 tell us more of the nature of the factors which go to 

 the making up of these pathological characters. 

 Many abnormal characters are doubtless more or less 

 complex in their nature. To treat such abnormalities 

 as simple units is but to disguise the truth. When 

 we know the nature of their complexities, the general 

 trend of the whole evidence leads us to believe that 

 they will fall into line with the general body of facts of 

 Mendelian inheritance. 



Some of the characters studied by Dr. Gossage 

 are of suggestive interest. In a certain family 



