PARNELL, AYYANGAR, RAMIAH, AND AYYANGAR 205 
The falling-off in weight of all three gold rice groups, as compared with 
the others, is very obvious. No explanation of this behaviour can be offered 
at present. In both parents the grain was perfectly normally developed. 
Purple rice. 
In a few, rather uncommon, varieties the colour of the rice is very dark 
purple. There is some variation in the depth of colour in different varieties, 
some being almost black (Plate V, fig. 3), and others rather lighter, with a 
reddish or brownish tinge on one side (Plate V, fig. 4). All the varieties seen 
with purple rice have been glutinous types, mostly from Burma. There is, 
however, no genetic connection between the purple colour and glutinous 
rice, 
Purple rice is dominant to white and gives a simple 3: / ratio of purple: 
white in F,. The total figures for a large number of families are given 
below. The purples are somewhat in excess but the figures undoubtedly 
represent a 3:7 ratio. 
Purple rice White rice 
54 families de af 17,850 5354 
Calculated 3 : 1 A. 6 T9509 : 5,852 
The main purple rice character is obviously due to a single factor. It 
appears, however, that there are other factors affecting the depth of the 
purple colouring. 
The F, of tho darkest type of purple x white, though showing some 
variation, is always lighter in colour than the purple parent, especially on 
the dorsal side at the lower end of the grain, and shows a distinct brownish 
tinge (Plate V, fig. 5). In F, the purple group shows considerable variation, 
ranging trom dark purple to a type where the purple is very considerably 
reduce] and a definite brownish tinge appears. It is quite impossible to 
separate distinct groups as the variation is practically continuous. 
In F, the darkest types breed true for purple though the very dark 
colour ot the purple parent is rarely reproduced, the majority being rather 
lighter and slightly brownish, resembling the lighter purple varieties 
mentioned above. The lighter types of F, purples are mostly heterozygous, 
showing the same type ot segregation as occurs in F,, though some distinctly 
light types have been found to breed true. 
There is no connection between the main purple rice factor P and the 
red rice factor R since white rice arises in F, from purple x red. F, is purple 
red since the red colour shows through where the purple is reduced (Plate V, 
