No. 156251. A thin fragment of matter consisting of minute roots of plants partially intermixed on one surface with sand. No. 156252. This consists almost wholly of plant rootlets and contains a very slight amount of sand. No. 156254. This consists of pieces of rotten wood through which had grown the rootlets of plants. The wood, upon a microscopical examination, is shown to be that of some dicotyledonous tree of a very loose and light texture. The plant rootlets in most cases followed the large ducts that run lengthwise through the pieces of wood and take up the greater part of the space. No. 156255. The mass contained in this bottle is made up of (_a_) grains, contained in their glumes or husks, of some grass, probably _Oryzopsis membranacea_; (_b_) what appears to be the minute spherical spore cases of some microscopical fungus. The spore cases have a wall with a shiny brown covering, or apparently with this covering worn off and exhibiting an interior white shell. Within this is a very large number of spherical spore-like bodies of a uniform size; (_c_) a few plant rootlets. No. 156256. The material in this bottle is similar to that in 156255 except that the amount of rootlets is greater, the grass seeds are of a darker color, seemingly somewhat more disorganized, and somewhat more slender in form, and that the spore cases seem to be entirely wanting. No. 156257. The material in this bottle is similar to that in No. 156249, containing the seeds numbered _a_, _b_, _c_, and _d_ mentioned under that number, besides a greater amount of plant rootlets and some fragments of corncob. No. 156258. This consists almost entirely of plant rootlets and sand. No. 156259. This consists chiefly of the leaves of some coniferous tree, either an _Abies_ or a _Pseudotsuga_. All the seeds with the exception of those of the leguminous plant are dead and their seed-coats rotten. The leguminous seeds are still hard and will be subjected to a germination test.[169] For a specific and positive identification of these seeds it will be necessary either for a botanist to visit the region from which they came or to have at his disposal a complete collection of the plants of the vicinity. Under such conditions he could by process of exclusion identify the seeds with an amount of labor almost infinitely less than would be required in their identification by other means. Very sincerely yours, FREDERICK V. COVILLE, _Botanist._ FOOTNOTES: