3. Yellow endosperm.

(a) More than one gene is involved in the difference of orange against yellow. In crosses between deep‑orange flint from the La Plata Region and Brazilian orange hard flint a bifactorial segregation of 15 orange to 1 yellow was found.

Ears	Total	Yellow	c^2
5	1592	107	0.60
6	2136	120	1.45

Descendants of the first lot gave in 20 ears normal F₂ and backcross ratios. In the next generation all ears again gave the expected ratios, but the proportion of non‑segregating, monofactorial and bifactorial ears was not in accord with expectation, as can be seen by the following data:

Descendants of (15:1)			Descendants of Backcross 3:1
	Ears			Ears
Segregation	Obtained	Expected	Segregation	Obtained	Expected
Non‑segregating	47	29	Bifactorial	30	27
Bifactorial	48	76	Monofactorial	51	54
Monofactorial	48	38
Descendants (3:1) Monofactorial			Descendants of Backcross 1:1
	Ears			Ears
Segregation	Obtained	Expected	Segregation	Obtained	Expected
Non‑segregating	38	34	Monofactorial segregating	44	44
Monofactorial	63	68

The frequency of ears with a monofactorial segregation is as expected, while that of ears with bifactorial segregation is abnormal. No explanation has been found at the present time and new plantings were made to check the results. The classification is very sharp. The recessive class being of a golden yellow, and the dominant one of orange shades, from deep to light. For the two pairs of factors, causing the difference between orange and yellow, we propose the symbols Or or and Or₂ or₂.

(b) In order to know how many genes are involved in the difference between orange and yellow, a special test was made last summer. Seeds of backcrossed ears segregating 1 orange : 1 yellow, of 11 different origins were planted, including different shades of yellow and orange. Crosses were then made in the following manner: every yellow was crossed with other yellows and with all heterozygous oranges. The orange types were crossed inter se. In order to check the dosage effect, every cross was made reciprocally, plant to plant, and at least two ears were secured of each cross. All were crossed with a y y tester. The data have not yet been fully analyzed but the results can be summarized as follows:

Cross	Segregations observed
Yellow x Yellow	All yellow	Orange and yellow	All orange
Orange x Yellow	--	Orange and yellow	All orange
Orange x Orange	--	Orange and yellow	All orange

This added one more proof to our hypothesis that more than one factor is responsible for the difference between orange and yellow. Some ears were planted this summer to test the segregation in F₂.

(c) We do not agree with some authors about the role of modifiers in differentiating orange and yellow. Our point of view is that the modifiers act only, as in other genetical segregations, in occasionally disturbing the ratios. We do not yet have enough data to confirm the hypothesis that a selected modifier complex is responsible for all the different shades of orange and yellow. Our working hypothesis is that several complementary factors are involved and proper tests must be synthesized for proving this assumption. As in the color of the aleurone layer, several basic genes are involved, besides dosage and modifier effects which cause a variation in shades.

N. Kobal