of the Genesee River, as well as a portion west of the Niagara, as its radius was one of fifty miles about Buffalo. The Cayuga

Flora of Professor Dudley (1886) was for the basin of Cayuga Lake and some adjoining ground, though covered in part already by Paine's Oneida list. The three lists mentioned do not record the plant. In the " Plants of Monroe County and Adjacent Territory," published in 1886 by the Rochester Academy of Sciences, it is listed for places near the Genesee River, being abundant in some of them. Macoun does not give it in any list of Canadian plants up to 1890, that being the date of some entries as "additions and corrections to parts I-IV" of his Canadian catalogue. I can add as a matter of personal observation, that in the summers of 1882 and 1884 I spent some weeks examining streams, lakes, and ponds in western New York for the study of Najadaceae, but collected other plants as well. The localities were principally south of the area recorded in the Rochester list and east of that of the Buffalo list, being in the counties of Wyoming, Genesee, Livingston, and some adjoining parts of Monroe and Ontario counties. I did not meet with the plant. Judging by the rate at which it has spread since it was first observed in the Desplaines valley, it is not likely to be present in a locality for any length of time without becoming abundant enough to attract attention, since it soon forms extensive mats or beds in favorable localities.

Though the year of discovery is not generally given in the publications cited, the time of publication is covered by ten years for places as widely separated as Rochester, Painesville, Detroit, and Chicago. This is about seventy years after the first notice by Barton and Nuttall by the Delaware. The distribution between these places and the seaboard and between one another, if in any way connected, must be ascribed to other causes than that of steady accretion of area along lines of natural or unaided seed distribution, however this may act in localities where a plant is once established. Nor are the places mentioned so connected by water communication that plants of this character would be likely to traverse the spaces in the reverse direction to the course of drainage, however this may aid when the direction. of flow is in their favor. Yet they are on main lines of railway

traffic, and to some extent of lake navigation, if these may have any connection with such seemingly sporadic dispersal of plants. That lines of railway are important factors in plant migration, especially for those of a weedy nature, is readily seen by one passing along their roadbeds. But there are evidently other means by which plants, whose seeds cannot be borne by currents of air, are able to cross widely intervening spaces. For those that grow in water or in the feeding places of migratory birds, seeds lodged in their feathers or in the mud that may cling to their feet is a plausible conjecture for dispersion. The transmission of undigested seed in the alimentary canal of birds is also the source of wide dispersion of plants. But when established, as in the case of this plant in the Desplaines valley, which has now been under observation nearly twenty years, the natural flow of the water bearing plants or seeds that may be taken up by it becomes a means of the more effectual dissemination in a given area. A specimen collected in 1892 by Dr. W. S. Moffatt on the banks of Salt Creek at Elmhurst has upon the label the statement: "abundant locally, covering several acres of creek-bottom." This being higher up the stream than where I found it in 1890, from its abundance may have been an earlier station and the source of those at Western Springs. Dr. Moffatt in the same connection mentions its presence at Riverside where Salt Creek enters the Desplaines River.

The case with the third crucifer, Sisymbrium altissimum L., is somewhat different, as it doubtless came into this region from the northwest; it is given as S. Sinapistrum Crantz in Macoun's Catalogue among the additions and corrections to parts I-IV, published in connection with part IV. It had then (1890) been "introduced in a number of places along the Canadian Pacific Railway." The earliest date recorded is 1883, at Castle Mountain, Rocky Mountains. In 1886 it is mentioned as by Lake Superior; in 1889 at a station forty-five miles east of Toronto. The first authentic record I have for Chicago is an unnamed specimen received from Dr. Moffatt, collected at Forest Glen, 1893; it was soon after seen by him in the western part of the city. The first place mentioned is on the Chicago, Milwaukee, and St.

Paul railroad and can well account for the line of introduction. It soon spread to various localities in and around Chicago, though I did not see it in the locality where I reside till 1900. In 1903 I found it common by the side of the Lake Shore and Michigan Southern railroad at Dune Park, Ind., thirty-five miles east of here. It is a quite common weed in the waste grounds of Chicago now. In Beal's Michigan Flora the first date given for a locality is Benton Harbor, 1896. This is on the east side of Lake Michigan, nearly opposite Chicago. The entry is also made, "later in many localities." As the Gray's New Manual states that it is "locally abundant as a pernicious weed" it may be considered as quite generally spread throughout the northern parts of the United States and the southern part of Canada. Since Britton and Brown give it a place as a ballast plant at New York, there may also be other centers of migration from eastern harbors, but the main line has evidently been from the northwest. The spreading of this weed has been quite rapid, gaining a large area in about twenty years. It produces seeds in great abundance. As I have observed it the height does not generally exceed 5 to 8 dm., that is, not very tall as one might infer from its specific name, though the stature is more or less influenced by the character of the soil. When crowded by its own kind or by other growths it may be very slender and but little branched, but with ample room it is bushy-branched, the diameter equaling or exceeding the height, or of a somewhat globular form, like a tumble-weed. Whether it actually functions as such I have seen no case, but the shape is one that suggests that it could be easily rolled by the wind if loosened from the ground by any means. These are the possibilities of a tumble-weed.

CHICAGO, ILLINOIS

ADDITIONS TO THE PLEISTOCENE FLORA OF NORTH CAROLINA*

BY EDWARD W. BERRY

In a previous paper the writer enumerated thirty-eight species. mostly forms which still exist, from the Pleistocene deposits of North Carolina.† Considerable new material, for the most part unstudied as yet, has since been obtained, from which the following have been selected for enumeration at the present time.

Juglandales

HICORIA AQUATICA (Michx. f.) Britton

Salix sp., Berry Journ. Geol. 15: 340. 1907.

Additional material has made possible the certain correlation of the specimen previously listed as a willow with this species of hickory. In the modern flora it is a denizen of low river banks and swamps from Virginia to Florida and westward in the Gulf region to Texas. It has not hitherto been found as a fossil, Station 850, Neuse River.

Fagales

QUERCUS MICHAUXII Nutt.

This occurrence is based upon fragments of leaves and characteristic acorn cups. In the modern flora it inhabits low, wet situations from Delaware to Florida and westward, but has not been previously obtained in the fossil state.

Station 850, Neuse River.

Ranales

LIRIODENDRON TULIPIFERA Linné

Berry, Amer. Nat. 41: 695. 1907.

Winged carpels of this species were recently recorded by the writer from the Pleistocene of Alabama, but leaves have not been previously recorded from American strata younger in age than the Cretaceous, although the genus is common in the Arctic and Eura

* Illustrated with the aid of the Catherine McManes fund. Berry. Journ. Geol. 15: 338-349. 1907.

sian Tertiary. The present record is based upon abundant and characteristic leaves collected by Dr. L. W. Stephenson from a clay lens in the sands of the Wicomico formation, one and onefourth miles east of Weldon. Fig. 1.

One of the specimens is shown in

FIG. I.

Liriodendron Tulipifera Linné, from the Pleistocene of North Carolina.

Rosales

CERCIS CANADENSIS Linné

Penhallow, Amer. Nat. 41: 446. 1907.

The accompanying figure (Fig. 2) shows a characteristic leaf of this species which comes from one and one-fourth miles east of Weldon. It has been previously recorded by Penhallow from the famous interglacial deposits of the Don valley near Toronto and in the modern flora it is said by both Britton and Small to range northward to southern Ontario. Both Sargent and Sudworth give its normal northern range as New Jersey and southern Michigan from which points it ranges southward to Florida and Mexico. It is essentially a warm temperate type, most of its near relatives being subtropical in habitat. Like the present species in this country Cercis siliquastrum Linné of southern Europe has been found in the interglacial deposits of France.