The communications in the forestry section of the Gardeners' Monthly for November, 1883, and January and March of the current year, relative to the growth of chestnut upon limestone soils, suggest the presentation of some notes under the above head.

That the nature and variety of the forest covering are to some extent determined by the physical and chemical constitution of the soil, sub-soils and the underlying rock formations, is well known to all geologists who have had much field work. So far is this fact recognized that in some cases the forest gives a clue to the geology. Granitic and hornblendic districts are distinguished by variations in the forest species. Limestone and sandstone, and the various calcareous and siliceous outcrops can be traced in their boundary lines by the forest. The slate rocks differ from the adjacent limestone belts. In the northern part of our country, or north of what is now recognized as the southern limit of the glacial phenomena, there is more or less mingling of rocks and earth in thick deposits, which cover the rock strata, and the mixed character of the forests is evident. But where this glacial covering is wanting and the soil is made up of the debris of the underlying rocks, as is the case in the country south of this old, continental moraine line, the forest bears a close relation to the rocky floor below, and the botany and geology are truly sister sciences.

For example, in the Highlands of New Jersey and in the South Mountain range of Pennsylvania, there are two well-marked types of gneissic rocks. And they are distinguished by their differences in the timber growing upon them. The feldspathic variety makes a light-colored, open, sandy soil, and the forest on it is largely chestnut and oak; the hornblendic rock makes a dark-colored, ferruginous and clayey soil, and oaks and hickory, with very little chestnut, make up the most of its forest. Descending into the limestone and slate districts of the great Kittatinny Valley, there is a noticeable lack of chestnut timber and a predominance of oaks (mainly white) and hickories. Black walnut also marks the limestone; on the slate this tree is comparatively rare. On the other hand, the wild cherry is common to the slate, but not to the limestone. The sandstones and the siliceous conglomerates are marked by more pine and rock oak with some chestnut. The various rock species which predominate in the broad, red sandstone belt of central New Jersey and southeastern part of Pennsylvania have produced sandy loam and clay soils, and the differences in the forests are here also recognized. One characteristic of the forest on the red, shaly areas is the almost entire absence of chestnut.

The trap-rock ridges which are in this red sandstone country bear more chestnut.

As Fuller, in his "Forest Tree Culture" says, the chestnut appears" to prefer a dry, sandy or gravelly soil to an alluvial, clayey, or very moist one." Bryant also correctly states it - when he writes in his "Forest Trees," that the chestnut "'seems to prefer the sides and neighborhood of hills and mountains, with a dry, sandy or gravelly soil." And, generally calcareous soils and the soils of limestone formations are not of this nature. The physical texture of the soil is apparently unsuited to this tree. But there are soils on some of the more highly crystalline, limestones, or marbles; as for example, some in Berkshire County, Mass., in Westchester County, New York, and in Northern New Jersey, which are largely made up of the fine rhomboidal fragments of this rock, and are, therefore, sufficiently open and dry to produce a luxuriant growth of chestnut. Much of the limestone of Chester County in Pennsylvania is of this nature. As a rule, the blue, sedimentary, limestone formations in all the Middle Atlantic States, cannot be said to be the home of the chestnut tree. Its most luxuriant growth and its largest size appear to be attained on our granitic and gneissic rock soils of the Appalachian chain.

New Brunswick, N. J.