20 



THE AMERICAN NATURALIST [Vol. LIII 



The results of the segregation into the three classes 

 gave 391 hulled, 661 intermediate and 426 hull-less. Here 

 the hull-less plants are too great in number, while the 

 number in the intermediate class is too small. The ex- 

 pected numbers are 369.5:739:369.5, with a probal^le 

 error of + 11.23. It is possible that some intermediate 

 plants were classed as hull-less. Such a condition is pos- 

 sible since some intermediates are found bearing only 

 one or two hulled kernels, and if these should be lost 

 through shattering, such plants would be classed as hull- 

 less when in reality they are intermediates. When the 

 grouping is made into the two groups, hulled and hull- 

 less, it is seen that the 3 to 1 ratio is approximated very 

 closely, as there are 391 hulled plants to 1087 hull-less, 

 giving a percentage of 26.45 + .76 hulled. 



An examination of this table shows further that some 

 of the families do not give ratios close to 1:2:1. This is 

 true with regard to certain families particularly with 

 certain of those coming from plants low in percentage 

 of hulled, and some of those relatively high in this re- 

 spect. The results of some of these families have been 

 brought together in Table IV. 



