100 This inheritance of characters in kiCE 



9 : o'o to : (J--/. Both factors nuist liave been segregating in tlie origii\al 

 9 : o'S family and tlie very wide deviation from a 9 : 7 ratio was no don)>t 

 due to the repulsion of both factors as already described. 



Eleven families, compiising 7,617 plants, have given such abnormal ratios, 

 ranging from .9 ; o*.S' to .9 ; G'4. These may be compared Avi'h the following 

 ratios calculated on the assumption that both factors are repelled to the 

 same degree : — 



This partial repulsion of ripening black factors may be compared with the 

 total repulsion, already described, of purple stigma and axil by the same 

 internode lining factor. It is of interest, in connexion with theories regarding 

 gametic reduplication in general, that, in certain cases, total repulsion 

 between purple lining and purjile stigma was recorded in the same families as 

 were giving the partial repulsion described above. 



(j) Eed rice. 



In the process of husking the inner glumes are removed from the padch/ 

 leaving the rice grain free. The colour of this unpolished rice varies in different 

 varieties from clear white, through various subdued shades of cream, yellow, 

 reddish-brown, etc., to the deep red of ordinary Q-ed rice (Plate V, fig, 1). A 

 still deeper shade, dark blackish -purple, also exists, but has been noted in only 

 one or two varieties of which crosses have not yet been examined. 



The colour is contained entirely in the pericarp which is largely removed 

 in the process of polishing to which the rice is submitted before being cooked. 

 Red rice is generally disliked as the polished rice still possesses streaks of red 

 colour in the furrows unless it is very thoroughly polished, in which case the 

 loss of weight and labour entailed are considerable. 



Many families segregating for the ordinary full-red colour have been 

 examined and, with the exception of a few cases to be considered later, this 

 character has proved to be a simple dominant, due to the presence of a single 

 factor, irrespective of any lighter colouring that may be present. 



The figures given in Table XIX indicate a very definite 3 : 1 ratio oif nil- 

 red to light colour. 



