Comparison of Open‐Pollinated Seed With Registered Seed of Three Cotton Varieties 1

This experiment was conducted near Athens, Ga., on Cecil sandy loam. The test area received a uniform fertilizer application of 240, 157, and 299 pounds per acre of N, P, and K, respectively, (360 pounds each of P,~O~ and KeO) each year, anal was irrigated at 30% available water in the top 2 feet of soil. New seed for the test was obtained each year from the Cotton Variety Test. This test consisted of 4 replications of 1-row plots 30 feet long with approximately 15 entries randomly placeal in each replication. Seed from Coker 100A, Empire WR, and *Plains’ rows was saved and all four replications of each were bulked to provide the open.pollinated seed the following year. The Cotton Variety Test area was relatively small and some crossing with all entries might have occurred. The three female parents of the open-pollinated seed were considered to be high yielding varieties well adapted to this area, but little is known of their combining ability. Coker 100A and Empire WR are known to possess some specific combining ability when crossed with each other3 (1), and Empire WR also exhibits specific combining ability when crossed with ’Pope’, one of the Cotton Variety Test entries in 1957. In the general area of the Cotton Variety Test each year, red leaf and normal green leaf strains were planted in alternate rows to determine the amount of natural crossing. By germinating seed from the green leaf rows cross pollination was estimated to be 45% in 1957, 35% in 1958, and 40% in 1959. These estimates are row to row crossing, total crossing within rows and between rows would be higher. Registered seed of each of the three varieties tested was obtained from an appropriate commercial source each year. The registered