i9io.] 



Mendelism. 



285 



such characters. Thus the Mendelian can predict what 

 will happen when, say, a white breed of rabbit is mated 

 with a coloured one ; he cannot predict the result of mating 

 a large animal with a small one ; he can foretell the colour 

 of the eyes of the children of two blue-eyed parents ; he is 

 ignorant of the law determining their height. Confining 

 our attention, then, to the inheritance of sharply defined 

 characters, of which colour will serve as a type, the root 

 principle of Mendelism may be simply stated. It is that 

 many, if not all, such characters behave as distinct units 

 in inheritance, and may be present (or absent) in the off- 

 spring, dissociated from the other characters present in either 

 of the parents, in accordance with certain definite numerical 

 laws. For example, a child may have the blue eyes of its 

 father, but all its other colour characters from its brown-eyed 

 mother ; moreover, the Mendelian law enables us to affirm that 

 the blue-eyed child has no dark-eyed character in its "blood," 

 even though its mother had dark eyes; in other words, the 

 offspring of this blue-eyed child, if mated with another blue- 

 eyed individual, will never show any " reversion" to dark 

 eyes. It cannot, however, be asserted that the offspring of 

 two dark-eyed parents will all have dark eyes, for it is a fact 

 that, whereas the blue eye is always "pure," in the sense 

 that it breeds true, the dark eye, on the other hand, is some- 

 times pure and sometimes impure, the "impurity" consisting 

 in the fact that the blue-eyed character is sometimes latent 

 and likely to appear in the offspring. In Mendelian termino- 

 logy, dark eye is "dominant" to the "recessive" light eye. 

 It must be clearly understood that dominance is not an 

 essential of the Mendelian law ; the root idea is that certain 

 characters are independent units, the transmission of these 

 units from parent to offspring being entirely indepen- 

 dent of the inheritance of other units which may distinguish 

 the parent individual. 



It is clear that we have here an entirely novel conception 

 I of heredity. The ideas hitherto prevalent, if capable of 

 ! definition at all, are associated with the use of the word 

 j "blood" in connection with heredity. It is assumed that, as 

 the child is of the same blood as its parent, it carries, in its 

 J constitution — it may be latently — something of all the 



