THE BLENDING OF PARENTAL TRAITS 195 



every species which is sexually dimorphic, possesses two kinds of 

 characters. In some characters parents resemble one another 

 closely ; that is, they differ normally only to the extent of fluctua- 

 tions. The inheritance of these characters is blended. In other 

 characters parents differ so widely that, did we not know them to 

 be male and female, we should suppose them to belong to different 

 species. The inheritance of these larger differences is apparently 

 alternative. Really it is blended ; only the reproduction is alterna- 

 tive ; for since parents transmit their characters through children 

 of the opposite sex to grandchildren of the same sex, we know 

 that, though every individual possesses only one set of non-sexual 

 characters, he possesses two sets of sexual characters, one of which 

 is patent and the other latent. If, then, similar characters blend 

 (e.g. the ears of one parent with the ears of the other, and so on), 

 there can be no doubt, in the case of the sexual traits, that the 

 patent male characters of the father blend with the latent male 

 characters of the mother, while the paten \. female characters of the 

 latter blend with the latent female characters of the former. The 

 child, however, presents a deceptive appearance of inheriting only 

 one set of characters in which there is no blending. Abnormally, 

 the patent sexual traits of the one sex blend with those of the 

 other as in many of the so-called hermaphrodites of sexually di- 

 morphic species. Except in the case of mutants, which in nature 

 only very rarely survive and have offspring, all the larger differ- 

 ences between mating individuals are sexual. But in some 

 characters mating individuals do not differ widely. Thus the 

 coloration of the male and female may be similar. The inherit- 

 ance of such characters is admittedly blended. In parthenogenesis 

 there is neither latency nor, of course, blending. Latency and 

 blending are phenomena of bi-parental reproduction only. 



322. The only real exception to blended inheritance that I can 

 think of occurs when a mutation first arises, If the mutant (the 

 individual that has mutated) be crossed with a normal individual 

 the mutation may be diminished, in which case the inheritance is 

 blended ; or it may be quite eliminated, leaving no representative 

 in the germ-plasm, in which case there occurs that extreme form 

 of ' blended ' inheritance which we have termed exclusive ; or the 

 mutation may be present in the offspring unchanged by conjuga- 

 exclusive inheritance is identical with gametic purity, but, as we have seen, it is 

 certain that gametic purity does not occur when ' inheritance ' is alternative. In 

 such cases we have only alternative development. The term prepotent as applied to 

 a parent, implies that his or her character predominates in the blend in any degree 

 up to an exclusive inheritance. To the Mendelian prepotence implies dominance. 



