PARNELL, RANGASWAMI AYYANGAR, AND RAMIAH 103 



It would seem from this that in these families the production of full-red 

 is due to the simultaneous presence of two factors, one of which, by itself, 

 produces the grey-brown type. In this case the three groups should show 

 0.9 : 3 : 4 ratio but the actual figures obtained are not very near to this. The 

 ratio of grey-brown to ivhite varies considerably in the different families and 

 for this reason it is hardly legitimate to total them together.* 



A red-riced plant, No. 617-N, from an F2 family that was segregating but 

 not counted, gave in F3 full-red 949 : white 753, a very definite 9 : 7 ratio. 

 Twenty-four segregating F4 families slxowed an extraordinary range of ratios, 

 moritly varying from 9 : o-3 — 9 : ■5-9 but with a few approximating to 9 : 7. 

 These ratios recall strongly the set of similarly varying ratios obtained in 

 connexion with the ripening black character, where there was evidence that 

 it was due to partial repulsion of the two determining factors by another 

 factor. It appears highly probable that the present case is due to a similar 

 phenomenon. 



Two plants in No. G17-N possessed slight reddish rice, quite distinct from 

 the full-reds, and proved to be natural crosses since their progeny contained 

 golden-glumed plants though this character was absent from the original 

 family. When the rice of these two famihes was examined it was found to 

 vary from definite slight reddish to pure white. This variation was so gradual, 

 as is usual in famihes segregating for this type of colouring, that definite groups 

 could not be separated, but in both cases all plants with golden glumes 

 possessed pure white rice. An exactly similar result was seen in a totally 

 unrelated family arising from a similar cross in another group. There can be 

 no doubt that in these three families some form of gametic reduplication was 

 taking place in which the factor or factors determining light reddish rice were 

 concerned. This fact supports the supposition that the varying ratios of the 

 families derived from No. 617 Avere due to gametic reduplication. 



The whole question will be carefully followed up, using selfed seed only 

 and noting all segregating characters in connexion with rice-colour. Attempts 

 will also be made to jiroduce red-riced plants by crossing white and grey-brown, 

 in the case of the 153 group, and different tvhite plants in the 617 group. Should 

 both the above types oi9 : 7 ratio be confirmed it would follow that red rice may 

 be due to the simultaneous presence of two factors of which at least one may 

 vary. 



With regard to the various light shades of red, brown, etc., it has proved 

 very difficult to obtain definite results. Many families undoubtedly showing 



* In the light of later evidence, see Postscript, this view is probably incorrect. [F. R. P.] 



