140 THE SCIENCE OF BIOLOGY 



ing segregation. The second might seem improbable, but 

 it can be accepted if it explains the facts. Suppose, there- 

 fore, that the Fi germ-cells do not carry B and W, each 

 diluted one-half, but B or W, and that the two kinds of 

 germ-cells thus produced are present in equal numbers. 

 The case would then be as shown by the figure. 



Suppose further, that each kind of germ-cell has equal 

 chances in fertilization. Referring to the diagonal lines 

 below the Fi germ-cells (Fig. 19), the germ-cell B may unite 

 in fertilization with another cell B or with a cell W; the cell 

 W may unite with a cell B or with a cell W. This exhausts 

 the theoretical possibilities and gives the ratio 1 B B : 2 

 B W : 1 W W, by the laws of chance. It is the same as in 

 matching heads and tails with a coin. If you throw heads 

 your opponent may throw heads, if you throw heads he may 

 throw tails, if you throw tails he may throw heads, if you 

 throw tails he may throw tails; which would be 1 H H : 2 

 H T : 1 T T A simple algebraic multiplication also illus- 

 trates the case: (x + y) X (x + y) = x 2 + 2xy + y 2 . The 

 fundamental ratio of 1:2:1 is thus explained as due to 

 chance combinations of germ-cells which are pure, in the 

 sense that they carry that which determines one member of 

 a contrasting pair of adult characters to the exclusion of the 

 other. Purity of the germ-cells with reference to the paired 

 unit-characters of the adult is thus a primary assumption in 

 the explanation of Mendelian ratios. 



The principles of Mendelian inheritance as thus disclosed 

 are: (1) that an adult animal possesses unit-characters, 

 which (2) segregate in heredity, and (3) that the determin- 

 ers for contrasting unit-characters are not carried together 

 in the same germ-cell. An adult organism is a mosaic of 

 these unit-characters, not a blend. Contrasting unit-char- 

 acters may be combined in the individuals of an Fi genera- 

 tion, but in subsequent generations they segregate in definite 

 ratios. The existence of the ratios is explained by assuming 

 that the germ-cells are thus pure with respect to a pair of 



