When infection occurred, the fractures failed to unite and, within the time limits of this experiment, the picture presented was essentially the same regardless of the type of fixation used. In clean wounds in Group I, progress of union was about the same with both types of plates, with a slight advantage in favor of the Eggers plate. Although accurate measurements were not used, we did not see evidence of the fragments sliding on the plates after the callus had formed. With clean wounds in the second part of the experiment, progress of the union was ahead on the side treated with the Eggers plate except in those animals in which the nail fitted snugly and in these it was in favor of the nail. We do not feel that the medullary nail delays union, but rather that lack of fixation due to a poorly fitting nail may delay union. Within the scope of this experiment it appears that fixation and apposition of the fractured fragments are more important in obtaininig union than the pressure of weight-bearing.