This investigation was conducted at Barrage Station (BHRS), Horticultural Research, Kalubia Governorate, Egypt, to get some new and promising pea lines (Pisum sativum L.) using pure line selection method on the basis of genetic variability. Selection was performed between or within four pea populations of Master Bcultivar. The populations were collected from four different regions in Egypt and subjected to selection procedure during 2012/2013, 2013/2014 and 2014/2015 growing seasons. The investigation intended to study different traits of growth and yield and its components viz., No. of days to flowering, No. of first flowering node, plant length, No. of branches per plant, pod length, pod width, pod thickness, pod weight, No. of seeds per pod, weight of seeds per pod, No. of pods per plant, weight of 100 seeds, shelling percentage, pod yield per plant and seed yield per plant. The target of this study was to improve pea yield and its components as well as earliness of yielding. The analysis of variance revealed highly significant differences among genotypes for most of the studied characteristics, indicating ample scope of selection. Most characteristics showed a higher heritability coupled with higher genetic advance through two cycles of selection indicating that these characteristics are stable and can be improved through selection based on phenotypic observations. Twenty lines were obtained from selection program and were compared against the baseline populations in addition to check cultivar (Entsar1).Results showed significant differences among lines for most studied traits. Selection program resulted in genotypes or lines given symbols of H3, G7, N1, N3, G2, G1 and K2 that considered promising lines.  These lines have good pods traits, higher productivity and earliness of flowering. Simple correlation coefficient analysis revealed that greater emphasis should be given to the traits of No. of branches and No. of pods per plant along with pod length, pod width, No. of seeds and weight of seeds per pod for yield improvement.

This work was conducted at the Experimental Farm of Nasser's   Faculty of Agricultural Sciences in Lahej Governorate, Yemen, during three   seasons 2003, 2004 and 2005. Five statistical procedures of relating yield components to yield; i.e., simple correlation coefficient, the path coefficient analysis, the stepwise regression, the multiple regressions and factor analysis were applied to seven yield contributing characters to determine their functional relationships to yield. Sixteen Maize genotypes were used in this study. Simple correlation coefficient revealed that, number of leaves/plant, ear height, ear length, number of rows/ear, number of kernels/row, 1000-kernel weight and shelling% had the greatest influence on grain yield/h. According to path analysis, weight of 1000-kernel had the greatest direct effect (22.23%) towards grain yield/h. While, number of kernels/row (9.33%) and ear length (9.32%) had the highest indirect effect to grain yield. Multiple linear regressions indicated that the variables which had the highest partial coefficient of determination in seed yield/h, were ear height, ear length, number of rows/ear and 1000-kernel weight (R2 = 43%, 22%, 9% and 12%, respectively). The stepwise regression shows that, 1000-kernel weight, number of kernels/row, number of rows/ear and shelling% were accepted variables which had the highest coefficients of determination with seed yield (88.9%). The factor analysis grouped 7 yield contributing characters in two factors, which altogether were responsible for 70.42% of the total variability in the dependence structure.