CORROBORATWNS OF MENDEL'S LAW 351 



caterpillars with yellow cocoons. The germ-cells of the hybrids 

 are, according to hypothesis, of four kinds, which may be repre- 

 sented by the letters (Y = yellow ; y = white ; G = striped ; 

 g = unstriped) YG, Yg, yG, yg. 



Now, the possible combinations of these in fertilisation are : 



YG with YG = YG = Yellow Striped 



Yg = YG = Yellow Striped 



t> ti yG = YG = Yellow Striped 



yg = YG = Yellow Striped 



Yg with YG = YG = Yellow Striped 



,, ,, Yg = Yg = Yellow Hsstriped 



yG = YG = Yellow Striped 



t yg = Yg = Yellow unstriped 



yG with YG = YG = Yellow Striped 



Yg = YG = Yellow Striped 



tt yG = yG = white Striped 



, . yg = yG = white Striped 



yg with YG = YG = Yellow Striped 

 Yg = Yg = Yellow unstriped 

 M ,, yG = yG = white Striped 

 yg = yg white unstriped 



9 Yellow Striped -f 3 Yellow unstriped -f 3 white Striped -f- I white 

 unstriped. 



Toyama's actual results show a very close approximation to 

 the theoretically to be expected results : 



Yellow Striped 6,383 individuals, 56-38% 

 Yellow unstriped 2,099 .. l8 '53% 



white Striped 2,147 . 18-96% 



white unstriped 691 6'i% 



The proportion 9 : 3 : 3 : i in 16 is called the normal Mendelian 

 ratio for a " dihybrid cross," where two pairs of contrasted 

 unit-characters are implicated. In each of the four groups 

 making up the 16 there is one individual homozygous, con- 

 taining units all similar viz. YG x YG (one of the 9 yellow 

 striped forms), Yg x Yg (one of the 3 yellow unstriped), yG x yG 

 (one of the white striped), and yg x yg (the single pure recessive 

 white unstriped). Any of these four, if mated with an individual 

 like itself, will breed true a point of great practical importance. 



