Abstract
The experiment include eight factorial treatments, represents the combinations between the two cotton varieties Lachata and SP8886 and four spaces between plants (15, 20, 25 and 30 cm), that carried out at Al-Hawyja, Karkuk Governorate during the season of 2007 using randomized complete block design with three replications. The seeds were planted through mid April using rows 60 cm apart. The data were taken for characters: plant height, number of fruiting branches, number of bolls per plant, boll weight, seed index, lint index and seed cotton yield. The results showed significant differences among varieties and planting spaces for all characters except number of fruiting branches, the differences among varieties for lint index were not significant. The interaction between varieties and spaces was significant for plant height, number of fruiting branches and boll weight. Trend analysis revealed that the regression relationships of plant spaces was linear with seed index, lint index and boll weight, and cubic with plant height, number of bolls per plant and seed cotton yield. It was shown that seed cotton yield had positive highly significant correlations with boll weight and seed index, and negative highly significant correlations with plant height and lint index. Path analysis results indicated that plant height and lint index had higher negative direct effects on seed cotton yield and important indirect effects through most of other characters, which indicates that these two characters could be used in breeding programs to improve yield, and stepwise regression analysis revealed that the best regression equation for seed cotton yield prediction was that which contained these two characters with determination coefficient equal 88.09%.