it any farther, holding it by the small end. It bent then. With a longer and stiffer pole, I could probably have fathomed thirty feet. It seems then that there is over this andromeda swamp a crust about three feet thick of sphagnum, andromeda calyculata and polifolia, and kalinia glauca, beneath which there is almost clear water, and under that an exceedingly thin mud. There can be no soil above the mud, and yet there are three or four larch trees three feet high or more between these holes, or over exactly the same water, and small spruce trees near by. For aught that appears, the swamp is as deep under the andromedas as in the middle. The two andromedas and the kalmia glauca may be more truly said to grow in water than in soil there. When the surface of a swamp shakes for a rod around you, you may conclude that it is a network of roots two or three feet thick resting on water or very thin mud. The surface of that swamp, composed in great part of sphagnum, is really floating. It evidently begins with sphagnum which floats on the surface of clear water, and accumulating, at length affords a basis for that large-seeded sedge(?), andromeda, etc. The filling up of a swamp then, in this case at least, is not the result of a deposition of vegetable matter washed into it, settling to the bottom, and leaving the surface clear, so filling it up from
