On the northern shore a large harbor, nearly 5 miles long and about 2 miles broad at its widest part, extends in a general westerly direction, ending in a vast waste of salt marshes interwoven with a network of creeks. Up this harbor the tides rush with great velocity, and when they sweep out to sea leave a broad expanse of flats, sandy on the north and central portions and muddy on the south. These flats cover an aggregate area of 400 acres, comprising 200 acres of hard sand and 150 acres of soft mud. Large stretches of these mud flats on the south are waste, and covered for the most part with eel grass. Other sections elsewhere are likewise waste for various causes, and are to be excluded as unprofitable or barren; yet the total available area remaining after making these deductions exceeds 350 acres. This is the theoretical condition,--the real condition is far otherwise: 20 acres at the most yield clams, and of these only 10 acres produce them in marketable quantities. The explanation of these conditions is interesting. In the winter the ice and the force of storms tear out great pieces of the tough marsh surf, and the tides sweep them down the harbor. Some of these huge masses are torn to pieces and washed away, others find lodgment on the broad surface of some tidal flat; these, becoming stationary, accumulate sediment; the grass grows upon them through the summer, and gradually a little island is formed. Surrounding these islands and oftentimes growing over their entire surface, bedded in among the roots of the marsh grass, we find a very thick set of clams. In short, all the digging of any kind is in the immediate vicinity of these islands. The deductions to be made from these facts are apparently simple. In the spawning season, when the microscopic clam larvæ are in their floating stage, they are carried here and there by the currents. Later, when they tend normally to settle in some fertile tract of flat, they are prevented from so doing by reason of the remarkable swiftness of the tides, which sweep strongly over the broad, smooth flats, and give the little clams no opportunity of lodgment. Only in the firm thatch of low-lying islands can they find anything to cling to, and here, with their slender byssus threads attached to unyielding grass or roots, they are able to withstand the wash of the current. Thus the clams are gathered in great numbers in these natural collectors, later are washed on the neighboring flat, and finally a little colony grows up about every island of this sort.