488 H. H. NEWMAN 



problem as to the factors involved in the relative developmental 

 success of reciprocal crosses, and of hybrids between various 

 species, is unsolved. 



4. Modes of inheritance in hybrids and the Mendelian hypotheses of 

 dominance and segregation 



It has been shown that structural and physiological characters 

 in the Fi generation of teleost hybrids show exclusive, blended 

 and particulate inheritance. There is seldom if ever pure domi- 

 nance and there is frequently an almost perfect intermediate 

 condition between parental types, especially in the case of pig- 

 ment characters. In some cases there is a sort of regional domi- 

 nance of one parental type in one area and a similar regional 

 dominance of the other parental type in another area; or else the 

 two types may exist side by side in quite intimate contact with 

 each other. These mosaics and mixtures are typical examples 

 of what we have come to call particulate inheritance. In the 

 case of integral variates the hybrids exhibit numbers of variates 

 intermediate between those of the parents. So we find practi- 

 cally all known types of inheritance in connection with pigment 

 characters alone, and would also find a similar condition for any 

 other group of characters studied equally intensively. The dis- 

 covery of the prevalence of intermediates and blends in the Fi 

 generation does no violence to the fundamental principles of 

 Mendelian inheritance, for it is practically certain that these 

 characters that seem to blend in the first generation of hybrids 

 would segregate out in tjrpical fashion in subsequent hybrid gen- 

 erations, were it possible to interbreed the Fi hybrids. We have 

 nothing to offer, therefore, that need be interpreted as out of 

 accord with recent Mendelian work, although previous work 

 dealing with the same kind of data has been cited not infre- 

 quently as furnishing examples of non-Mendelian inheritance. 



