12: 192), by Dr. S. E. Jelliffe in 1899 (Fl. L. I., xi-xii), and by Dr. J. W. Harshberger in 1904 (Proc. Acad. Nat. Sci. Phila. 56: 605).

The pine-barrens of Long Island are almost confined to the southern half of Suffolk County, but extend a few miles westward into Nassau, mostly in the shape of isolated patches. Dr. W. C. Braislin (Abstr. Proc. Linn. Soc. N. Y. 17–19 (1904– 1907) 94. O 1907) places "the western limit of the scrub pines" at Central Park, about three miles west of the county line, where there are perhaps three or four hundred acres of pinebarrens; but I have seen other unmistakable patches of the same vegetation two or three miles farther west, namely, between Hicksville and Westbury and also about half way between Merrick and Hicksville.

The pine-barren area corresponds approximately with that of the soil mapped as "Norfolk coarse sandy loam" in the soil survey report on western Long Island by J. A. Bonsteel and others (Field Operations of the U. S. Bureau of Soils for 1903). Like most other unglaciated portions of Long Island, it is almost perfectly flat, with a barely perceptible southward slope of about 15 feet to the mile. A few very shallow valleys containing brooks or creeks traverse it, generally from north to south. It lies entirely south of the newer or Harbor Hill moraine, but partly north of the older or Ronkonkoma moraine. The soil seems to be of the Columbia formation, one of the youngest of coastal plain deposits. Its chief constituents here are silt and coarse sand. Just why pines should predominate on the “Norfolk coarse sandy loam" and deciduous trees on some equally sandy soils similarly situated a little farther west is not clear. It cannot be altogether a matter of water-content, for the pines occur also well within the edges of the swamps.

In Babylon and Islip, the two westernmost townships on the south side of Suffolk County, pine-barrens are the prevailing type of vegetation, and as there are a good many square miles of them in these townships entirely uninhabited, they are in excellent condition for study. My observations on the flora of Babylon and Islip have been confined to three trips on foot

across the pine-barrens from north to south in the fall of 1907. On October 6 I traversed the pine-barrens of Islip in going from Smithtown to Babylon by way of Brentwood and Edgewood; five days later I went from Hicksville to Babylon (station) by way of Pinelawn, and on November 3 from Cold Spring station to Amityville, the latter half of this journey being also through the township of Babylon.

There seem to be only two natural types of vegetation in the region under consideration, namely, dry pine-barrens and swamps. The swamps are confined to narrow belts along the streams, and the only one I have examined is that of Santapogue Creek, two

FIGURE I.

Dry pine barrens about one-half mile south of Edgewood station (town of Islip), Oct. 6, 1907. Trees all Pinus rigida, the largest about 9 inches in diameter and 30 feet tall. Undergrowth almost entirely Quercus ilicifolia and Q. prinoides in equal proportions, about 4 feet tall. The picture embraces a horizontal angle of about 36°. This scene is typical of thousands of acres in both townships.

or three miles west of Babylon station. The dry pine-barrens are fairly uniform over many square miles, the principal natural. variation being that toward the northern edge the shrubs are

smaller and the herbs more numerous than elsewhere. Southward the arborescent species of oaks become larger and more numerous and gradually crowd out the pines. The northern boundary of the pine-barrens seems to be more sharply defined. The average appearance of the dry pine-barrens is illustrated better by the accompanying photographs* than it could be by any description. The pines are the dominant feature of the landscape, and the underbrush consists chiefly of a dense growth of two shrubby oaks,† all the individuals of both reaching approximately a uniform height in any one locality. The herbs are more numerous in species, but more scattered and inconspicuous. The dry pine-barrens have probably always been subject to occasional fires, which since the advent of civilized man have become frequent enough to kill a good many of the pines but otherwise have perhaps caused little change.

The species observed in this habitat in the two towns mentioned are as follows. They are divided first into trees, shrubs, and herbs, and then arranged approximately in order of abundance in each class.

* None of the existing local floras of Long Island is illustrated, so far as known to the writer, so these may be the first photographs of Long Island pine-barren vegetation ever published.

† These two oaks are remarkably similar in appearance, considering that they belong to different sections of the genus. See in this connection Rehder, Rhodora 9: 61. 1907.

Vaccinium spp.

Smilax glauca

Rhus copallina
Arctostaphylos Uva-ursi

Chrysopsis Mariana
Aster concolor

Epilobium angustifolium
Gaultheria procumbens.

Besides these, Aster spectabilis, Laciniaria scariosa, Lespedeza capitata, and Sarothra gentianoides were seen along some of the roads, and they may perhaps also occur naturally in the pinebarrens. Chrysopsis falcata, which is commonly regarded as a typical northern pine-barren plant, I have found only on a gravelly hill in Smithtown,* and (abundantly) in gravel between the rails

At left,

FIGURE 2. Two scenes in the pine-barrens of Babylon, Oct. 11, 1907. dry pine-barrens about a mile south of Pinelawn. Pines growing more densely than usual. Oaks mostly Q. ilicifolia, four or five feet tall. At right, east edge of swamp of Santapogue Creek, looking south, just above the road from Farmingdale to Babylon. Shows principally Pinus, Acer, and Betula.

of an old railroad which runs eastward from Garden City, Nassau County. Hudsonia ericoides, another supposed pine-barren plant, I have seen only on a high gravelly hill in the southwestern part

* Most of the stations cited for it in Jelliffe's Flora of Long Island are on the north side of the island, among the hills.

of the town of Huntington, just north of Babylon. Corema Conradii may possibly occur somewhere in this region, for according to Mr. J. H. Redfield* it is associated, wherever it grows, with a good many of the species listed above.

The flora of the swamps is considerably richer than that of an equal area of dry pine-barrens. The single pine-barren swamp examined on October II contained the following species, some on its edges, some right in the stream, and some in intermediate positions.

Chamaecyparis thyoides, which grows in some more or less similar swamps in Nassau County, ‡ and Polygala lutea, which is

* Bull. Torrey Club 11: 97-101. 1884.

† Described in Britton's Manual 70. 1901. Formerly referred to the more southern A. glomeratus (Walt.) B.S.P. (A. macrourus Michx.).

See Torreya 7: 198-200. O 1907.