6 POLLINATION AND CROSS-FEKTILISATION IN RICE, 



such cases were found, and in each case as many red-grained seeds 

 as could be picked out were sown and the seedHngs transphmted 

 separately in 11)12, every precaution being taken to prevent 

 accidental mixture. In three of the six plots the red seeds have 

 given rise to a few white-grained plants, proving that these were 

 heterozygous as regards seed colour. 



To return to the variant plants mentioned above, seed of 

 fuur of these, which it was suspected might be crosses, was care- 

 fully preserved, sown and transplanted separately in 1912, and all 

 have civen rise to a variety of types as described below : — 



The first of these (labelled Ba 1), was found in a variety 

 named Gobrabali, a coarse rice fi'om Bakarganj, and differed from 

 the type chiefly in having reddish-coloured awns. This in 1912 

 split into four types — 



(1) Ketl awns, twu iiuier glumes ;it Howeiiug time pale green 



with red tip, turning jellow when ripe, stigmaa black ... <iU 



(2) Red awns, inner glumes at Howering time dark green with 



red tip, turning mottled brown, stigmas black ... 13 



(3) Wliite awns, inner glumes at flowering time pale green with 



nu tip, turning yellow, stigmas white ... ... 1!' 



(4) White awiis, inner glumes at (lowering time dark green with 



no tij), turning brown, .stigmas white ... .. 4 



Itatio red awns to white awn.s, 317: 1 



The .second (labelled K 20) was found in a \ariety named 

 Ukulmadhu froin Kajshahi. The parent plant had a small, fine 

 grain, the two small outer glumes yellow, and the inner glumes 

 yellow with brownish coloured apex. This in 1912 split into 

 three nuiin ty[)e.s 



(I) Small outer glume.s vellow, inner glumes at fir&t green with 

 reddish tip, turning yellow with reddish tip, stigmas 

 purple ... ... .. ... 5 iHited. 



{•2) hitto, but with .stigma.x white ... ... ... 5 „ 



(3) Small outer glun>e8 coloured reddish-brown, inner glumes at 



fust nddi.'^h with no tip, turning blai k, stigm.is white ... 2 



The only obvious difference between types (1) and (2) in this 

 ca.se lay in the colour of the stigma which can he accurately de- 

 teiiiiinfMl niily Ijefore pollination has taken place. rnfortunatelv 



