The purpose of this experiment was to determine the effect of dividing the periosteum circumferentially on the growth of long bones and the effect of partial division of the periosteum on the formation of angular deformity of growing bones and to investigate the tethering effect of the periosteum on the epiphyseal plate as its possible mechanism. One hundred twenty rabbits, one-month old with average weight of 570g, were divided into four groups. In the first group, the periosteum of the right distal femur was completely divided circumfer- entially; in the control group, the medial aspect of the right distal femur was surgically approached but the periosteum was not divided, in the third group, the medial half of the peristeum of the right distal femur was divided transversely; in the fourth group, the medial half of the periosteum was divided and resutured. Following the operation, X-rays were taken every two weeks until twelfth week and then at twentieth week and the length of the femur and the physis-shaft angle were measured. l. After complete circumferential division of the periosteum, the ratios of right over left femoral length before operation, and two, four, six, eight, ten, twelve, and twenty weeks postoperatively, were 100.0±0.0%, 102.1±0.1%, 102.±0.0%, 102.9±0.0%, 103.5±0.0%, 103.6±0.0%, 103.5±0.0%, and 103.5±0.0%, respectively. The overgrowth was most conspicuous during the first two weeks. On the other hand, in the control group there was slight overgrowth, but this was not statistically significant. 2. After partial division of the periosteum, the physisvhaft angles before operation, and two, four, six, eight, ten, twelve, and twenty weeks postoperatively, were 91.0±1.2, 87.8±2.5,84.4±2.4, 83.1±3.2, 84.2±3.5, 86.4±2.3, 86.2±2.3, and 86.8±1.9, respectively. The valgus deformity was induced most conspicuously during the first four weeks, whereas partial correction of the de- formity took place between the eighth and tenth week postoperatively. After repair of the partially divided periosteum, only 0.9 of angular deformity was observed at sixth week, but this became statistically insignificant at tenth week as were in the complete division and control groups. 3. Histologically, no difference was observed in the cartiage cells of the epiphyseal plate and the bony trabeculae of the metaphysis between the control group and other groups. The divided periosteum regenerated at the ninth week, this being same also in the repaired group. There was no statistical difference in the thickness of the epiphyseal plate between the operated femur and the opposite normal femur in the completely divided group, this being same in the control group. Threr was also no statistical difference in the thickness of the epiphyseal plate between the medical and lateral halves in the partially divided group, this being same in the repaired group. Based on these results, the investigator was able to draw following conclusions. 1. In the growing long bone, complete circumferential division of the periosteum near the epiphyseal plate results in stimulation of growth and lengthening of bone. 2. Partial transverse division of the periosteum results in asymmetrical overgrowth, inducing an angular deformity. 3. It is suggested that the mechanism of overgrowth following division of the periosteum lies in the removal of the tethering effect of the periosteum. These findings may bear clinical implications on leg length inequality and angular deformities following certain fractures and epiphyseal injuries during growth period.