Previous autografting and homografting experiments (Iten and Bryant, '75; Stocum, '75b) showed that, when progressively later stages of upper arm blastemas were transplanted to either the distal radio-ulna or the wrist, surviving grafts of each stage formed wrist and hand parts in 100% of the cases. However, the earliest stages transplanted formed the humerus and radio-ulna at a low frequency, with an increase in this frequency when later stages were transplanted. These results can be interpreted in two ways: (1) The proximal-distal organization of the blastema is stable at the earliest stages with the low frequency of humeral and radio-ulnar formation being attributable to artifacts introduced by surgical trauma, artifacts which would not have as much effect on later stages; (2) The proximal-distal organization of the early blastema is initially labile and can interact with the stump to conform to the level of transplantation. This lability would decrease in later stages. The present study describes the results of new experiments which argue against the latter alternative. In addition, differences between autograft and homograft development are described which could be attributable to an immune response directed at homografts.