PAENELL, RANGASWAMI AYYAXGAR, AND RAMIAH 



Table XII. 

 Type B, ly.B, 23. 



93 



In the six Fo families, totalling 2,129 plants, no case occurred in which 

 L and G were not coupled. Ii\ the twenty-three splitting F3 families, totalling 

 28,281 plants, nineteen showed failure of this couplii\g, as shown in itahcs 

 in Table XII. 



Unfortunately the Fo parents of these families were not selfed and, there- 

 fore, it is not absolutely certain that these abe:rrant t}^es were not crosses 

 from outside the group. Crossing inside the group could o)\ly slightly affect 

 the 1:2:1 ratio but could not produce the aberrant types. 



A few undoubted cases of crossing from outside were found, but these 

 were easily recognizable since the ij^Q of internode and glume colouring was 

 quite distinct from that of the group, which was peculiar to itself. 



The nineteen aberrant plants were absolutely typical of the group in all 

 respects and there appears to be little doubt that they represent examples of 

 incomplete couphng of L and G. It \\\\\ be seen from tables that follow that 

 occasional aberrant plants of these tv-pes are constantly occurring in families 

 segregating for L and G and this fact supports the idea of incomplete coupling. 

 To make quite certain of this point, and in order accurately to estimate the 

 amount of coupling, fifty plants of the LGSA type have now been selfed. 

 In view of this no further reference to this point will bo made beyond giving 

 the numbers of such aberrant types as occur in the famiUes described. 



No departure from complete coupling occurred with regard to purple 

 stigma and axil. 



