escarpment and the present margin of the lake is a series of parallel terraces, each showing the altitude at which the receding waters have made a pause. These terraces along the southern shore of Lake Erie range from ninety to one hundred and twenty feet above the present level of the water. In Huron and Sandusky Counties the subsidence of the escarpment permits them to diverge a greater distance from the lake. Curving northward into Michigan, they pass through Monroe, Wayne, and Oakland Counties, and continue in that direction. They are recognized again on the shores of Lake Michigan.

At Mackinac Island are deeply engraved records of a higher level of the waters. The island itself is but a vestige of an ancient formation which once filled the straits, and joined the highlands on the west and south. It towers, a monument of the destructive agencies of geogony, three hundred and eighteen feet above the lake. The principal plateau of the island is one hundred and fifty feet above the lake. In the midst of this rises "Sugar Loaf," one

hundred and thirtyfour feet higher (Fig. 85). This is a remnant of the rocks forming the upper plateau which once extended over the whole island. It is a conical mass of brecciated limestone, all whose sides bear the marks of the longcontinued action of the waves. On one side an ancient grotto

Fig. 85. View of Sugar Loaf, Mackinac Island. has thus been exca

vated, into which the surges have rolled with the deafening reverberations of a sea-coast "purgatory." The principal plateau of the island is limited at nearly all points by an abrupt wall dropping down into the deep waters of the lake. Here is a beetling cliff, one hundred and forty-five feet high, called "Lover's Leap," connected with a senti mental Indian legend. In another place is "Chimney Rock," one hundred and thirty-one feet high, and in another, "Robinson's Folly." At "Arched Rock" (Fig. 86),

on the eastern side, towering one hundred and forty feet above the lake, the fierce waves, unable to reach the solid and unyielding brow of the precipice, have mined beneath it, perforating the limestone wall; and a natural bridge hangs there, with one end resting on a winged abutment stretching toward the lake. All round the walls of this castellated and charming isl and the recording waves have left their hieroglyphs, from the water's edge to the battle.

LOSSING-BARRICT.

ments, and he who can read the language may ponder there the vicissitudes of the ages.

While, during the high tides of the lakes, the erosive waves were gnawing at the rocks of Mackinac and Ohio, the waters of Lake Michigan, in a quieter mood, were performing a work equally enduring and peculiar. Here we find our attention challenged by the question of prairie origin and prairie features, but the views to be presented will be held in abeyance until a chapter on a subsidiary topic shall have been interposed.

L 2

VENTURE here to enunciate a view which to many

may appear incredible. For some years past I have been inclined to believe that the germs of vegetation which flourished upon our continent previous to the reign of ice, and many of which must have been buried from twenty to one hundred feet beneath the surface of the glacial rubbish, may have retained their vitality for thousands of years, or even to the present time. There are not a few indications that vegetable germs are capable of such preservation, and not a few that they thus exist in the ancient drift. The consequences of such remarkable preservation possess a geological importance so novel and interesting that I am sure the reader will be pleased with a view of the facts bearing upon the doctrine.

Many familiar facts may be cited which certainly have a significance far greater than has been generally suspected, and which tend to show that the seeds of vegetation are reposing in a dormant state in our ordinary soils and subsoils. Nothing is a more common observation than to see plants making their appearance in situations where the same species was previously unknown, or for a long time. unknown, and under circumstances such that the supposition of a recent distribution of seeds is quite precluded.

The sudden appearance of unwonted species frequently occurs when a change is produced in the physical condition of the soil. Left to Nature, certain perennial grasses. secure almost exclusive foothold in our fields, and form a

sod in which the ordinary annuals are unable to flourish. Break up the sod after any number of years, and subdue the perennial grasses, and we shall have a crop of annuals the first season-Veronicas, Chenopodiums, Euphorbias, Portulacas, Ambrosias, Crab-grasses, Foxtails, Panicums, etc. Cease cultivation, and the Poas and Glycerias will immediately resume possession. Similarly, the pertinacity with which the common knot-grass seizes and maintains its position only along the hardest-beaten footpaths is notorious, while the greater plantain renders itself no less conspicuous growing alongside. Earth thrown out of cellars and wells is generally known to send up a ready crop of weeds, and, not unfrequently, of species previously unknown in that spot. In all these cases, after allowing for all known possibilities of the distribution of seeds by winds, birds, and waters, it still seems probable that germs must have previously existed in the soil.

Similar sudden appearances of new forms take place when a change is effected in the chemical nature of the soil. Illustrations are familiar to every agriculturist. How soon does a dressing of undecomposed muck, or peat, or sawdust develop a crop of acid-loving sorrel, and how readily is it again repressed by a dressing of some alkaline manure? Let the waters of a brine-well saturate a meadow, and how long before we witness the appearance of the maritime Scirpus and Triglochin, or some other saltloving plant whose germs, unless spontaneously developed, must have lain dormant at a greater or less depth?

Something of the same nature is witnessed on the disappearance of dominant species, whether through the agency of man or the processes of Nature. It is well known that the clearing of a piece of forest and the burning of the brush is almost always followed by the appearance of certain unwonted plants known as "fire-weeds." In many