H. Drinkwater 329 



in the case of two types of sweet-pea, and his i-esults have been 

 repeatedly confirmed by other observers. The normals (recessives) 

 should produce normals only : the abnormals (dominants) should pro- 

 duce normals and abnormals in about equal numbers. The chart at 

 once shows that the types produced accord with this theory. What 

 about the relative proportion of the two types among the children 

 of the abnormals ? The short-fingered children of abnormal parents 

 ought to be equal or very nearly equal in number to the normal- 

 fingered children of the same parents. 



As a matter of fact this equality is not usually exact, and theory 

 does not require it to be so, for there is a chance element which comes 

 into play which makes this precise number uncertain and variable. 

 According to Mendel the germ cells of any individual, such as one of 

 these abnormals, are of two kinds. One kind of cell carries the factor 

 which can produce the abnormality, and is inherited from the abnormal 

 parent ; the other kind of cell, being inherited from the normal parent, 

 lacks this factor : and these two kinds of germ cells are present in the 

 individual in equal numbers. If this be true, it must of necessity 

 follow that the particular kind of germ cell which will take part in 

 any given fertilization, will depend, so far as we can tell, upion a chance 

 meeting ; just as there is a chance of drawing a black or a white marble 

 out of a hat which contains an equal number of each. It is a lottery 

 with equal chances for both kinds. Sometimes one kind will pre- 

 dominate, sometimes the other, but in the long run the numbers 

 will be apjjroximately, if not exactly, equal. 



In this family there are 50 abnormals, so that according to Mendel's 

 theory there ought to be ahoid 50 normals. The actual number of 

 normals is 48. 



Thus, instead of the exact 50 per cent, of abnormals which miglit 

 occur in strict accordance with theory, we have 51 '02 per cent. This 

 is sufficiently near to constitute this family a striking instance of 

 Mendelian inheritance in the human subject. This remark applies 

 also to the other instances of Brachydactyly published by Farabee and 

 mj'self, as well as to my illustrations of Minor-Brachydactyly. 



If a normal member of this family were to ask whether if married 

 to another normal (relative or non-relative) there would be any risk 

 of having Brachydactylous children, one would be justified in replying 

 that there was no risk whatever of such children resulting from the 

 union, for, not in a single case has a short-fingered child been bom of 

 two normal parents. This abnormality thus differs in a remarkable 



