IN THE TROPICS. 
379 
to change the green colour to yellow. This change does not take 
place uniformly, and so a piebald effect is often produced. The 
original sample of seeds was almost entirely green. It was kept in 
a stoppered glass bottle exposed to the light of an ordinary room, 
and by the end of six months the cotyledons on the exposed side 
had all become completely yellow. Other kinds of peas with green 
cotyledons showed a similar change, but this took place more slowly 
than in the case of Telephone. 
A majority of the first crop seeds at Peradeniya were more or less 
yellow when gathered, distinctly more so even than in the case 
of Telegraph. They were separated into three groups—green, 
piebald, and yellow, respectively—the groups being as it happened 
nearly of equal size, and 50 seeds of each kind were sown. The 
crop from each group was again separated into a like threegr oups, 
but the offspring of yellow seeds did not include a larger proportion 
of yellows than the others. The exact numbers of these samples 
are of no value, owing to the want of definite limits to the different 
groups, but it was clear that the tendency to turn yellow was inherited 
nearly equally by the offspring of all three groups. In the above 
cases the seeds were left on the plants for some days after all were 
ripe ; cross-bred plants, on the other hand, were always taken up 
before the seeds in the ‘topmost pods were quite hard, and the seeds 
examined at once ; under such circumstances the green colour could 
almost always be detected in the parental strain also. 
B .—The Cross-bred Forms. 
1. Native pea No. 1 x Telephone. —Twenty-one cross fertiliza¬ 
tions were made, of which only 9 were successful. The 9 pods 
yielded 20 seeds, none of which differed at all from the normal seeds 
of No. 1. 
The crosses were made in November, 1902, and the seeds were 
sown immediately after gathering ; all but two germinated, but only 
14 plants produced seeds. The plants reached a height of 3-4 feet, 
nearly the same as that of the best plants of the two parental forms 
sown at the same time. In the character of their foliage and in 
stoutness of stem they resembled Telephone, but there appeared to 
