one or more of them come in contact with the upper portion of the female cell they blend with it and their contents are absorbed and fertilization is effected. Soon a rapid growth of filaments and cells takes place at the base of the female organ as a result of this fertilization. In fact there is formed a naked cluster of spores (c) from these filaments, all fertilized by the single sexual act upon the central female cell.

In these few pages the endeavor has only been to point out a few of the leading methods of asexual and sexual reproduction among fresh-water alga, and we feel in closing that the vast subject has been but here and there touched upon. But enough has been said to show that even in these lowly forms the too often supposed sameness of reproduction loses itself in variety of methods and multiplicity of changes.

SURFACE GEOLOGY OF THE MERRIMACK VALLEY.1

BY WARREN UPHAM.

THE highest fountains of Merrimack River are Eagle Lakes, on Mt. Lafayette, 1090 feet below its summit and 4170 above the sea. The source of the straight river is a lake which lies in the deep Franconia Notch, beneath the jutting rocks of the Profile. This stream is at first inclosed by high mountain ranges, and descends more than 1200 feet in its first nine miles. Distances and heights along this river are as follows: Profile Lake, about 1950 feet above the sea; mouth of East Branch, 9 miles, 710; at Plymouth, 28 miles, 468; at New Hampton, 39 miles, 438; mouth of Smith's River, two miles below Bristol, 45 miles, 320; mouth of Winnipiseogee River at Franklin, 55 miles, 269; mouth of Contoocook River at Fisherville, 66 miles, 249; mouth of Soucook River, 76 miles, 199; Amoskeag Falls at Manchester, 89 miles, 179 to 123; at line between New Hampshire and Massachusetts, 108 miles, 90; Pawtucket Falls dam, Lowell, 117 miles, 87; Essex Company's dam, Lawrence, 128 miles, 39. The entire length of this river is about 155 miles, and its last twenty miles are affected by the tide.

The Merrimack Valley in New Hampshire is comparatively straight, and forms a continuous line of depression which is a principal feature in the topography of the State. Its course is

This essay is principally based upon explorations made for the Geological Survey of New Hampshire, and will be more fully presented in vol. iii. of the report on that survey.

slightly east of south. The upper and lower portions of the river which occupies this valley are known by different names. For more than fifty miles from its source, this river is called Pemigewasset; and the name Merrimack is applied to it only from the confluence of the Winnipiseogee River with the Pemigewasset at Franklin.

After entering Massachusetts the river turns to the east at North Chelmsford, and thence pursues a devious east and northeast course at right angles to its valley in New Hampshire. It here threads its way among hills, with no distinct, wide valley; and only low water-sheds divide it from adjoining basins on the south.

In considering the surface geology of Merrimack Valley, we will begin at its head and describe first the modified drift which forms conspicuous plains, terraces, and intervals or bottom-lands along its course in New Hampshire, and occurs in gravel ridges, similar to the Kames of Scotland, well shown at many places along the whole course of this river; next, the prominent rounded hills of coarse glacial drift or till, which are finely displayed along this river in Massachusetts; and, last, the marshes and beaches at its mouth. After this, we will inquire what these deposits teach in regard to the history of this valley in the glacial age, during the melting of the great northern icesheet, and since that time.

The modified drift of the upper part of the Merrimack, called Pemigewasset River, is usually one half mile to one mile wide, and is bordered on both sides by high hills or mountains. Below Franklin the modified drift is usually one to two miles wide; its greatest development is in Concord, in Merrimack, and in Litchfield, where it has a width of nearly four miles. The hills which border this part of the valley rise with comparatively gentle slopes, and the lowest points of its eastern water-shed are only three hundred and fifty to six hundred and fifty feet above the

sea.

On Pemigewasset River modified drift occurs first in Lincoln, five miles from Profile Lake. This is very coarse, water-worn gravel, containing pebbles six inches to one and a half feet in diameter, or sometimes larger. It has an irregularly smoothed surface, imperfectly terraced, with its outer margin twenty feet above the stream. From this point modified drift is continuous on one or both sides of the river for thirty miles. In the first seven miles, to Woodstock village, it consists wholly of gravel of

A

different degrees of coarseness. Southward, banks and terraces of sand begin to appear, but gravel still predominates for a long distance below. The stream here frequently occupies a broad, shallow channel paved with pebbles of all sizes up to two feet in diameter, with little admixture of fine gravel or sand, which accumulates only in deep or sheltered places.

For ten miles south from the mouth of East Branch, or nearly to the south line of Thornton, a high terrace of gravel or sand is commonly well shown on both sides of the river, and has a uniform, continuous slope of fifteen feet to the mile. This slope is nearly the same as the descent of the river, which has evidently swept away this deposit to a depth of from seventy to one hundred feet over the area occupied by its channel and bordering bottom-land or interval. Nowhere else in New Hampshire is the erosion of the modified drift, by which it has been shaped in terraces, so clearly displayed. Here it seems certain that a former flood-plain, ten miles long, has been terraced as we see it by the excavation of the river.

In Campton the Pemigewasset receives two considerable tributaries from the east, Mad and Beebe rivers, which drain basins on the northwest and southeast of the mountain range that culminates in Sandwich Dome. South of the Beebe River the upper terrace, increased in height by alluvium from the tributary, forms a pine-covered plain, one mile long and a half mile wide. These "pine plains," appearing in a few places on the Pemigewasset and commonly along the Merrimack, form one of the characteristic features of this valley.

In Plymouth and Holderness both the high plain and interval are finely shown, and the extent of the alluvial area, at one point a mile and a half wide, is greater than at any other place on Pemigewasset River.

Dunes. In the north part of New Hampton and in many places for thirty miles southward to the north line of Concord, we find numerous dunes or sand-drifts lying at various heights on the east side of the valley up to three hundred feet above the highest terraces. These dunes appear in large amount and reach their greatest height near their beginning, two miles south of Ashland. Here the sand-drifts, one to five feet deep, are strewn in a pathway ten to twenty rods wide, which extends a quarter of a mile along the hill-side, with a northwest-southeast course, rising three hundred feet above the ordinary modified drift, or to a height about eight hundred and fifty feet above the sea.

These dunes occur only on the east side of the valley, consist wholly of fine sand, and lie in trains which ascend from the highest terrace in a southeast direction along the hill-side. All these characteristics indicate that they owe their origin to the transportation of sand, by the prevailing northeasterly winds, from the plains below, probably at the period when these had their greatest extent, prior to their excavation by the river, and, we may presume, before the appearance of a forest. They are usually made conspicuous at the present time by being blown in drifts, which are so constantly changing that they give no foothold to vegetation; but when they occur at considerable heights the lower portion of the series is generally grassed over, making the upper drifts appear isolated on the hill-side. The whole train of dunes before mentioned is equal by estimate to a mass one thousand feet long, fifty feet wide, and two feet deep, thus containing one hundred thousand cubic feet or five thousand tons, which has been raised by the wind an average height of one hundred and fifty feet.

Another very good illustration of this transporting power of the wind is found in Sanbornton, one mile southeast from Hill. Here, as also in New Hampton, the ancient dunes have been swept forward anew since the land was cleared. The sand from a hollow one hundred and fifty feet long, forty feet wide, and two to five feet deep has been carried, in long northwest-southeast drifts, two hundred to four hundred feet farther and twenty-five to thirty feet higher up the hill. The depth of recent excavation is shown by a large stump which has been thus undermined.

Similar dunes, high above the ordinary modified drift, occur along the east side of Connecticut River in New Hampshire and southeast from Ossipee Lake.

From New Hampton to Bristol the river flows westerly, almost at right angles with its general direction, descending by a nearly continuous slope eighty-six feet in the four miles, this being the most rapid portion of its course south of East Branch. Here it is closely bordered by sloping hills, and differs from all the rest of this valley in New Hampshire in being well-nigh destitute of modified drift. The high terraces reappear below Bristol, and thence to Franklin have a height one hundred and fifty to one hundred and seventy-five feet above the river.

From Franklin to the Massachusetts line the ancient high flood-plain of the Merrimack is everywhere well shown by conspicuous terraces. Along much of the way these terraces expand

on one or both sides into wide sandy "pine plains," so called because their principal wood-growth consists of white or pitch pines. These are often accompanied by a thick and tangled undergrowth of scrub oaks, which, with the pitch pines, flourish best on these barren plains. The terraces have a very level surface, with a regular but slight slope, which amounts to nearly the same as the descent of the river.

At Franklin the upper terrace is well defined upon both sides of the valley. It has here considerable fall in a short distance, being four hundred and forty-five and four hundred and forty feet above the sea at the north side of Webster Brook and Winnipiseogee River, and descending in less than a mile to four hundred and thirty and four hundred and twenty at their south side. In the next nine miles the upper terrace falls to a height one hundred and twenty-five feet above the river, which continues for more than twenty miles to the north part of Manchester, the highest terrace seeming to descend most rapidly near the present falls of the river, so that a nearly uniform height above the river is maintained.

In Canterbury the upper terrace spreads out into plains which are at some places a mile wide. The Boston, Concord, and Montreal railroad through this town is upon these high plains, while the Northern railroad in Boscawen and Concord lies on the lowest terrace, being embanked much of the way to raise it above the floods of spring. The plains of the south part of Canterbury, extending one mile into Concord, show an unusually rapid, continuous slope, amounting to eighty feet in four miles, or from one hundred and thirty to only fifty feet above the river. The north end of this slope appears to be at the normal height, representing the level of the river at the time of deposition of these plains, while its south end is about seventy feet below this normal line, which is here shown on the west side in the plains north and south of Fisherville.

Boscawen village is built on the south end of a similarly sloping terrace, three miles long, in which distance it falls thirty feet, and we find thirty feet more fall of the same terrace in less than a mile along the village street. The whole of this terrace is below the normal height, showing a deficiency of fifteen feet at its beginning and of forty feet at the north end of Boscawen village.

The supply of alluvium brought down by the river at this point was not sufficient to fill the valley to its average depth.