FOR the sake of chronological consistency, as well as for the intrinsic interest of the subject, this volume should begin with some account of prehistoric Andover. The accompanying plates,¹ prepared to illustrate a single geological feature, reveal at the same time both the topographical peculiarities of the town and its relation to surrounding political divisions. The territorial centre of "Old Andover," before the division of the town,² lies twenty-one miles north and four miles west of the State House, in Boston, in latitude 42^o 40', north, and longitude 5^o 54' east of Washington. The Merrimack River bounds it upon the north, while the Shawshin River, rising in Lincoln and Lexington, passes diagonally through the town from southwest to northeast. The southeast portion of the town is drained by tributaries to the Ipswich River. Hagget's Pond and Cochichawick, or Great Pond at North Andover, form distinct drainage basins, and empty into the Merrimack by separate outlets.

1 These were drawn by Mr. G.W.W. Dove. 
For larger size more detailed images of all three - 1751KB or 614KB or 
Plate 1 - 194KB , 569KB; Plate 2 - 181KB, 514KB; Plate 3 - 225KB, 651KB. 
The above three plates are 64KB, 60KB, 75KB respectively, clicking on each map brings up the middle size image of about 200KB.

2 Andover, before its division, contained about one sixth of the territory of Essex County. Its boundaries, before a portion was set off to Lawrence in 1847, were, the Merrimack River on the northwest (separating from Dracut and Methuen), Bradford and Boxford on the northeast, Middleton on the southeast, Reading and Wilmington on the south, and Tewksbury on the southwest. In 1709 the town was divided into two parishes, North and South, from the latterof which the West Parish was set off, and incorporated in March, 1827. In 1855 the town was divided nearly by the parish lines, the north division being incorporated as the town of North Andover. The name of Andover was relinquished to the south division, with whose institutions of learning it had become almost essentially identified. This transfer of the name makes it difficult to separate the history of the two towns, as the present North Andover was,, for more than two hundred years, Andover.

The  "tablets of stone," containing the geological history of Andover, are so mutilated that we can read only the first chapter of the record and the last. The rocks are devoid of fossils, and belong to the Laurentian formation. This part of the story may well be styled sketches of " Old Andover," for the rocks are, according to Prof. C. H. Hitchcock, among the very oldest to be found in the world, antedating by a vast period the strata of the White Mountains of New Hampshire. The prevailing coarse gneiss rock, appearing near the depot in Andover, contains a large amount of iron, and upon exposure to the air rapidly disintegrates and becomes unsightly. Through the Centre of the town, cropping out at Rattlesnake Ledge near Foster's Pond, at Sunset Rock, and upon the grounds of the Theological Seminary, there is an intercalation of granite about a half mile wide, remarkable for the size of its crystals of feldspar and for its freedom from iron. This terminates on the north at Carmel Hill, its eastern boundary being near Prospect Hill. To the southeast of this point there are extensive beds of an impure soapstone -a magnesian rock allied to the dark hornblendic rocks, which occur both in that vicinity and in several other portions. of. the town. The extreme southeastern portion of the town is crossed by the same uplift of ancient dioritic rocks (parallel to the general direction of the Appalachian chain), which contains the Newburyport silver mines. A belt of mica slate crosses the northeastern corner of the town.

But during the long period in which the Devonian and Carboniferous and Tertiary rocks of the central and western portions of our continent were being deposited, Andover was surmounted by an elevated plateau of land, which by its denudation was furnishing the gravel and sand and clay, out of which these later sedimentary deposits have been formed. The rocks which we here see exposed are but the stubs of mountains, which, during countless ages of exposure to denuding agencies, have been worn down to their present level.

While the geological record of Andover, during the earliest periods, is largely an untranslatable hieroglyphic, and the middle portion is absent, we are compensated by the abundance and intelligibility of the later record. The marks of. the glacial epoch in Andover are open to inspection before every man's door. Glacial striae can be seen, among other places, upon the rocks beside the road, a half mile out from Andover towards North Andover ; on those back of the Punchard School-house; on the exposed quartz crystals of Sunset Rock, and in the vicinity of the school-house,in Scotland district. Excellent exposures of glaciated rocks appear on the old turnpike from Lawrence to Salem, just beyond the North Andover line in Middleton, also on the Salem road from Andover. Scratched stones also abound everywhere in the " hard pan " at various depths.

Prospect Hill, rising four hundred and twenty-three feet above the sea, is one of the highest points of land in the county, and belongs to a very remarkable class of elevations, connected with the glacial period. Wood Hill, Pole Hill, Claypit Hill, Boston Hill, Woodchuck Hill, and the whole series of hills extending through to Great Pond, and surrounding it, are not, as might be expected, rocky elevations, but are, vast heaps of unstratified compact clay containing scratched pebbles and gravel, and littered over with angular boulders.¹  The distribution of this class of hills over so much territory, as is represented in the plates, is shown, on No. Ill. These elevations have been named by Prof. C. H. Hitchcock " lenticular hills,". from their peculiar lens-shaped outline, as seen upon the distant horizon. This series of hills continues to the northeast as far as Portsmouth, N. H., and in an irregular course may be traced westward to the Connecticut River. A remarkable cluster of them appears also in the vicinity of Boston, upon one of which the State House stands. The hills in Charlestown and Chelsea, and in Boston harbor, are also conspicuous examples. So also is Asylum Hill in Danvers, and numerous others, extending to Ipswich Neck. Abelt of land, four or five miles wide, from which they are absent, separates this series of lenticular hills from that passing through Andover.

1 These boulders and pebbles have all been transported from the north. A well known belt of porphyritic gneiss cropping out in the neighborhood of Weirs, N. H., about eighty miles to the northwest, has furnished Andover with numerous unmistakable specimens. The pebbles of mica slate are perhaps from localities nearer by.

The best explanation which can be given of these unique and to the geologist perplexing hills, is that tbeyare the remnants of an old terminal moraine, roughly marking what was for a long period of time the southern border of an earlier glacier in New England. Subsequently, upon the extension of this ice border to the south shore, this earlier moraine was first covered up beneath the ice sheet, and then by the movement of the ice over it was partially broken tip and sculptured into its present forms; The general trend of this series of hills is northeast and southwest, but the longer axis of the individual hills is usually from northwest to southeast, which is the direction of the ice movement, as shown by the scratches upon the rocks, The extreme terminal moraine of the continental ice-sheet, composed of similar material to, these hills, but continuous, forms the backbone of Cape Cod, of the Elizabeth Islands, of Long Island, and Staten Island, appearing at Perth Amboy, in New Jersey, and crossing that State to Belvidere, in Pennsylvania.

A later glacial deposit (now known in scientific circles as Kames) is represented in Andover by such formations as " Indian Ridge,"¹a portion of which is shown in detail upon Plate 1. Kame is a Scotch word, meaning sharp ridge. The extension ofthe system through the town is seen in Plate II., and of this and another series, through the county, in Plate III.

1 " The Great Ridg " is the term used in deeds one hundred and fifty years ago.
2 The measurements were made under superintendence of the writer by the class of 1875, Phillips Academy.

These ridges are composed of clay, sand, gravel, and pebbles of all sizes, up to those which are four or five feet in diameter. In most places there are some signs of irregular stratification. But frequently for a depth of twenty feet or more, signs of stratification are entirely absent. The stones in this formation are never scratched as in the hard-pan ; but the are all more or less subangular, showing abrasion of some kind. ' These too are largely from the north.

In the ". Transactions of the Association of American Geologists and Naturalists," for 1841 and 1842, Pres. Edward Hitchcock, of Amherst College, gave a detailed account of Indian Ridge, so far as then observed.¹ He there characterizes it as "decidedly the most interesting and instructive case [of the kind] which he had met with." A map of a mile and a half of it, then supposed to be its limit, was given by President Hitchcock in the same paper, prepared by.Prof. Alonzo Gray, of Phillips Academy. This map, on a reduced scale, reappears in " Hitchco&'s Elementary Geology,"² and covers nearly the ground of our first plate. Some other ridges of a similar nature were noticed by him, and the suggestion was made that further, researches might show a system where now onlya confused group was observed.

1 See page 198.
2 See page 260 (30th edition).

. xxii   PREHISTORIC ANDOVER.

We could not improve upon the description of the main features of this formation given by Dr. Hitchcock in 1842.

" Our moraines form ridges and hills of almost every possible shape. It is not common to find straight ridges for a considerable distance. But the most common and most remarkable aspect assumed by these elevations is that of a collection of tortuous ridges, and rounded, and even conical, hills with corresponding depressions between them. These depressions are not valleys, which might have been produced by running water, but mere holes, not unfrequently occupied by a pond."

In 1874, the writer ascertained that this belt of ridges extended through the whole length of the town of Andover, as shown in Plate II., striking the Merrimack at the upper end of Lawrence, passing a little west of Frye village, crossing the Shawshin at Ballardvale, and forming the shores of Foster's Pond. Subsequently, in 1875 the details, as then ascertained, were published in a Bulletin of the Essex County Institute, at Salem, showing the extension of the system south as far as Wakefield, and north, to the New Hampshire line. The direction of the belt of ridges is northwest by southeast, conforming nearly to that of the glacial striac. Karnes frequently pass over the lenticular hills where their height is less than two hundred feet, and descend into shallow depressions, crossing river valleys without ceremony. Still later investigations brought to light a parallel belt of gravel ridges, reaching the sea at Beverly, and continuing north through Topsfield, Boxford, and Haverhill far into New Hampshire. These two series of kames are shown in Plate III.

I When once the clew was discovered, numerous parallel systems of kames,were found, stretching back in many instances from near the sea to the base of the Mountains. In passing from Andover to New Brunswick by inland routes the traveller crosses more than thirty kames, each of which is as imposing as the series with which we are familiar in Andover, and some of which are continuous for one hundred and twenty miles; making the map of the kames and moraines of New England look like a gridiron.2 Besides those marked upon our map, six or seven other kames cross the. Merrimack Valley between Newburyport and the ang1e of the river, in Tyngsborough. These are all, however, less clearly defined and more subject to interruptions than the Andover or Haverhill series.
 

1 Transactions ofAmericau Association of Geologists and Naturalists, for 1841 and 1842, P. 191.

2 Prof. Geo. H. Stone, of Kent's Hill, Maine, is the authority upon this subject in that State.

PREHISTORIC ANDOVER. xxiii

A kame system comes down from Hudson, N.H., through Tyng's Pond in Tyngsborough, Mass., and passes
through Chelmsford. Another crosses about three miles above Lowell. Another just above the Pawtucket Bridge in Lowell,
and on the other side of the river appears at the Poor Farm. Three miles below Lowell is still another very clear instance
of a kame's crossing the river valley, where it is for some distance covered with alluvium. This kame -appears south, in.
Tewksbury, towards Long Pond. Another is seen west of. Hagget's Pond, appearing also one half mile cast of the State,
Almshouse and again in North Woburn. The kames east, of Hagget's Pond and those which match them north of the
river, appear to be merged farther south in the main Andover series. A mile below Lawrence again, upon the north side
of the river, is a small kame in line with the gravel deposits running through North Andover, and appearing as a kame
near Marble Ridge Station, and farther south in Middleton. A well-marked kame, also, comes down from Amesbury to
West Newbury.

The manner in which kames (called in Sweden Asar) were formed, has long been a source of contention among geologists. It is possible that the early settlers supposed kames to be, like many mounds at the west, the work of Indians, and hence their name, "Indian ridge," not only in Andover, but in other places. More probably the name arose from Indians' choosing the kames for camping grounds, or burial places, as the whites now frequently do. No one who has studied them carefully, and is aware of their extent, could suppose the kames to be artificial. The most probable theory of the origin of these remarkable ridges is that they, are somewhat of the character of medial moraines, and mark. the, courses of the surface' flow of water during the last stages, of the melting ice sheet. The ice had doubtless been thousands of feet in depth, and when the material forming the kames was deposited, still filled most of the depressions, and lingered in such transverse valleys as that which the Merrimack follows in the lower part of its course. Superficial streams, swollen by the action of the summer sun, would at that period, flow _with great violence during the hot season, and their course would be marked by vast accumulations of coarse gravel, which would in some places be lodged in ice channels, in others spread out over masses of ice. Finally, as the last masses and the lowest stratum of ice melted, the gravels thereon would settle down from the ice (as dirt does from snow- drifts in the spring) into the irregular forms in which we find these ridges.

Hagget's Pond doubtless marks a depression where the ice lingered while a kame-stream deposited in a temporary lake the sand plains to the south towards Tewksbury. Pomp's Pond was preserved from filling up by a similar mass of ice. The " kettle holes " near Pomp's Pond, and in the plain at Ballardvale, mark places where smaller masses of ice were covered tip by the sand and gravel. When the ice melted, a hole would be formed without any outlet. The basin of Great Pond in North Andover was formed in a different manner. In this case the lake is hemmed in by lenticular hills, one of which partially dams its natural outlet. Lenticular hills have also in many places below North Andover determined the course of the Merrimack River.

I Thus it appears that the citizen of Andover does not need to go to Switzerland, nor to Greenland, to study glacial phenomena. But he may enjoy that privilege to his heart's content among his own hills and gravel deposits. A most instructive portion of the skeleton of a continental ice sheet is spread out before his own doors. It is in gravels contemporary, in the period of their deposition, with the formation of Indian Ridge, that the pal~eolithic implements of northwestern Europe and eastern North America, especially in Trenton, N. J., are found. It is not improbable that the peat bogs, and "kettle holes," and ponds of Andover may furnish material aid in determining the antiquity of the glacial age, and so, of man, in America.' G. F. W.