210 



R. RUGGLES GATES. 



these ratios makes it evident that they nearly all fall remarkably 

 close to three or four ratios. So close is the fit that it seems 

 probable that several ratios, such as 5:1, are significant as such, 

 though at present no complete explanation of them can be of- 

 fered. I was formerly inclined to regard some of these ratios 

 as the expression of merely quantitative differences without 



TABLE II. 



Fz (Oe. grandi flora X rubricalyx}. 



2 (Oe. rubricalyx X grandiflora). 



1 By "impossible" is meant that the chances against this interpretation, taken 

 in connection with the other results, are so great that for practical purposes it 

 need not be considered. 



