Abstract
1. The leaf tensile strength of individual plants of a population of Lolium (multiflorum x perenne) perenne was measured on four occasions, October and December 1963 (160 plants) and January and March 1964 (73 of the 160 plants). All plants had been chosen for the good uniform vegetative growth they had exhibited over several seasons. The cellulose content of the 14 weakest and 14 strongest plants in October and of all 73 plants in March was determined.2. There was a wide range in leaf strength, maintained from October to March, within the population. In October all 14 weakest plants had lower cellulose content than the 14 strongest and from these two groups six plants of low leaf strength and low cellulose content and six of high leaf strength and high cellulose content were selected as parent plants for divergent selections. In March there was a positive highly significant (P < 0.001) phenotypic correlation between leaf strength and cellulose content (r = +0.58).3. Two diallel crosses were performed, one within the six parents of low leaf strength and low cellulose content and the other within the six high parents. Leaf strength and cellulose data was obtained from progeny and from parental ramets in October 1964 and high overall heritabilities for both characters (approx. 0.8) calculated from achieved advances and parent/progeny regressions. Selection for high leaf strength and high cellulose content was more effective than for low.