No. 550] 



With relationships expressed as 

 ciation, probable errors can be calc 

 comes possible to determine whether 

 between observed ratios and those 

 different gametic formulas are close 

 from chance. 1 



The difficulty of securing a 7 : 1 di: 

 omous cell division was appreciated 

 collaborators in the example first cil 

 that the coupling was in an 8: 1 rath 

 and kept in mind for a time until a gi 

 imated that resulting from a gameti 

 ratio was obtained. Kegarding the 

 7:1 and 8:1 ratios to explain the r 

 the first experiment, the authors : 

 Saunders and Punnett, 1908, p. 3) : 



But this indecision was apparently < 

 quent paragraph on the same pag( 

 examples that were referred to a IS 

 that the couplings follow a definite 

 in the following statement : 



