Florida Agricultural Experiment Station corrected and the trees are given every facility for making nor- mal growth. Soil and seasonal conditions are always changing, so it sometimes happens that the conditions under which the disease is developing become corrected without any effort on the part of the grower, and the trees grow out of the diseased condition. This may happen where Sthe disease has developed following a drouthy 1 condition in the soil, or from some partic- S ular overfeeding of the trees which is not repeated. :It has already been stated that it usually requires more than one flush of growth for the disease to develop. It also requires more than one flush of growth, even with treat- ment, for the trees to grow out of the dis- ease. Neither the preventive nor the curative methods remove the symptoms already devel- oped in the tree; both merely prevent their further development in succeeding flushes of growth. Other conditions being equal, it is probable that some of the curative treat- ments induce an earlier prevention of the de- velopment of the gum symptoms than the preventive treatments. Among the different preventive practices carried out in the grove to cure dieback, starving the trees for ammonia and stopping cultivation entirely, or reducing it to a min- imum, are the ones more commonly em- played. The effectiveness of these methods depends much upon the soil conditions. FIG. 10. Bark ex- Ammonia Starvation.-Where it is clear- crescences on a sub- terminal branch ly evident that the disease is due to overfeed- ing with the organic ammoniates, ammonia starvation will be helpful. The extent of starvation necessary will vary. In the average grove ammonia fertilizer is applied in three applications each year. The entire omission of ammonia from the fertilizer is a dangerous practice. If the ammonia is omitted entirely it is safest to omit it from one application only. In succeeding applications a low percentage of ammonia should be used and this practice not continued beyond two or three