INHERITANCE OF DIGITAL MALFORMATIONS IN MAN. 73 



of a short fingered parent has short fingers; and that no long 

 fingered descendant of a short fingered parent ever had short 

 fingered children. There is no historic evidence, so far as 1 can 

 learn, to support the first part of the tradition ; it may, or may 

 not, be true. The fact that there is a tradition concerning 

 the anomaly, without accounting for its origin, may be taken 

 as partial proof that the origin is so remote that it has been 

 forgotten. The second part that exogamy has been the custom, 

 is true for at least five generations, as will be seen in table V. 

 It would be very interesting indeed if this part of the tradition 

 should be violated . 



HEREDITY. 



Probably the most important part of this study is that 

 relating to the remaining portions of the tradition concerning 

 heredity. At present the question of heredity is one of live 

 interest on account of the testing of Mendel's discovery, — the 

 law of heredity. The present case demonstrates the fact that 

 the law operates in man as well as in plants and the lower ani- 

 mals. The abnormality here is shown to be the dominant 

 character. The tradition that every other child has short 

 fingers, '' is not quite true; yet, as nearly as possible, half the 

 offspring have the anomaly. This is in perfect conformity 

 with the law, the underlying principle of which is the purity 

 of germ-cells and their production in equal members. When 

 there is a union of normal and abnormal individuals, the abnor- 

 mals producing germ-cells N and A in equal numbers, the 

 chances are equal that germ-cell N of one sex may unite with 

 germ-cell N of the opposite sex, or that germ-cell A of one sex 

 may unite with germ cell N of the opposite sex. Since the 

 abnormal character is shown to be dominant, the chances are 

 even that the offspring may be normal or abnormal. Accord- 

 ing to the laws of chance we should not expect that every other 

 child would be abnormal, as in the tradition, but we should 

 expect the total number of normals and abnormals in a large 

 series to be very nearly equal, and that is what we find to be 

 true here. 



