The study of several hundred thin sections from most of the magnetite mines and prospects in the Adirondacks over a period of twenty years, has convinced me that the ore is in large measure due to high temperature metasomatic replacement. James F. Kemp in 1897 suggested this mode of origin. But his early paper has been forgotten or ignored by most of the later investigators, although both Colony and I have called attention to it. As the ore itself is largely the result of replacement, the method of attack was obviously through a study of the wall rocks. This shows that a series of overlapping processes, involving differentiation, assimilation, crystallization, contact action, deuteric, metasomatic, pegmatitic and hydrothermal introductions, took place. Gallagher has proved that most of these processes operated at Lyon Mountain. This paper extends these into a generalization applicable to most of the magnetite deposits of the Adirondacks.