In this study, using autoradiographic techniques, our attention was primarily directed to the question whether or not, and under which conditions, the scion’s labelled photosynthates would pass into the rootstock. Until recently, in the absence of stock leaves the melon/ ' Cucurbita graft appeared to be completely incompatible. In a specially devised system of water culture, this turned out not to be true, however, since stock-defoliated grafts after an extremely critical period are able to recover gradually but completely. From then on they show a remarkably good growth. Microorganisms attacking the roots are taken to be responsible for recovery not taking place if the grafts are grown in soil. Trying to account for this apparent incompatibility and subsequent recovery, the problem should be approached from different angles. The one way of approach concentrates upon the degree of perfection of the graft union, in the anatomical sense. The other major line of thought is biochemical in nature, and considers the question how far the biochemical functions of stock and scion are complementary, or balanced. For the present, it seems justified to consider the graft union to be anatomically perfect. The “temporary incompatibility”, then, should be due to the absence of a “translocation factor”. The subsequent recovery can be visualized to be due to a process of biochemical adaptation. Looking upon a graft union as a mere mechanical barrier in translocation studies, is to be critisized, the more so since still other factors may be involved, such as the degree to which the various parts of the graft act as “sources” and “sinks” with regard to each other.