Rough computations of our observations were made in the intervals between the working seasons, and these acquainted us with the progress we were making, and pointed out the direction which succeeding inquiries should take. From the results of our labors we gained at each step confidence and encouragement. What appeared at first a tangled skein of accidental or inconstant causes, we ultimately recognised as orderly and harmonious relations; and, our methods of observing improving steadily, the work advanced to its close at a pace constantly accelerated.

The observations of the past season were confined to no special locality, but were made at various points where previous operations were incomplete or required connecting links.

Our field work commenced the first of June, and the quict weather which prevailed during this month was improved for the occupation of the more exposed stations-those near shallow portions of the bar and along the outside coast. We had designed to occupy a station which should, if possible, lie quite beyond the reach of the New-York harbor drift, and enable us to determine whether any oceanic current sweeps into the great bay formed by the coasts of Long Island and New-Jersey. For this purpose we anchored, in thirty fathoms water, nearly sixty miles east-southeast of Sandy Hook, where, during a period of fifteen hours, we measured the currents at the surface, and at depths of twenty-three and one hundred and fifty feet, besides a few determinations of the mean motion for the entire depth. At this station, nearly forty miles from the nearest land, we found regular tidal currents, nearly as strong as those observed at the light-ship the previous season. No oceanic current could be detected, but the augmentation of the ebb current, caused by the drainage of the land waters, was very appreciable. The velocities of the currents are not so regular at this station, from the fact that the depth of the moving water stratum is variable, at one time extending to the bed of the sea, at another reaching but a short distance below the surface. The directions of the flood and ebb drifts were found to be respectively west-southwest and east by south; which, making due allowance for the disturbing effects of the land waters, would indicate that the tide-wave has here a westerly motion. The land waters of which we have spoken are doubtless the combined drainage from New-York harbor and the various inlets; for, extending our observations along the south shore of Long Island, we found that they outlive the tidal currents, and establish themselves as a constant coastwise stream along the eastern portion of Fire Island beach.

The stations outside of the bar were eleven in number, at which above seventeen hundred observations were recorded, and of these more than five hundred were made at points below the surface. The greater part of our season's work lay in the lower bay or in the vicinity of the bar, where there remained some localities unexamined, and others at which previous examinations had given discordant results.

From the computations which followed the field-work of 1858, it appeared that where observations were sufficiently numerous the causes of a certain class of shoals were immediately deducible from the data obtained. It was ascertained, on making a composition of the currents at each station, with the assumption that they are to be regarded as forces acting simultaneously, that the resultants take directions towards the shoals as focal points; making it evident that the sand which forms these shoals

is gradually swept together from the neighboring channels. Simple as the dynamics of this natural process may be, its form can only be developed from the most accurate determinations of the elements. The resultant, for instance, may be a very small quantity from a station at which the adverse currents are very violent. In a case like this, the slightest error of observation, or even the selection of an unsuitable period, may give us a false result and lead us entirely astray. If the observations are not sufficiently frequent, they may fail to give the exact durations of certain phases of the currents; or if the positions of the stations are not closely determined, errors enter into the directions of the forces. Again: If the observations are not continued long enough to eliminate the diurnal inequalities, an undue weight will be given to some of the elements which enter into the problem. At the commencement of the past season, forewarned of these difficulties, I placed in the hands of my observers printed rules for their guidance, and required of each person a strict conformity to them. Twenty-one stations were occupied in the portion of the work to which I have just referred, and at these the aggregate number of observations reaches nearly five thousand, of which above eighteen hundred are from points below the surface. At these stations the observations were usually kept up in unbroken series of twenty-five hours each.

A more suitable period for observations of so exact a character could scarcely have been chosen; our operations were rarely suspended by bad weather, and few delays of any kind occurred.

In making observations upon bars and shoals, the disturbing effects of strong winds cannot be disregarded; for it not unfrequently happens that they change the direction of the current, or wholly reverse its course. In districts of shallow water the waves created by the winds have a motion of translation whose effect upon the log is very great; and although the observer is able to distinguish this sudden and uncertain motion from that of the more steady current, he cannot introduce a correction for it. Where the sea is deep, the impulses it receives from the winds result in simple undulations, giving to the log no horizontal motion whatever, so that, even when the swell is very heavy, accurate current observations are possible. As far as our experience has gone, we have never observed in the waves any power of transportation where the depth of water exceeds three fathoms.

Above the Narrows there were eight stations occupied-three in the main channel of the harbor, two in the Hudson and three in the East River. At these there were recorded over seventeen hundred observations, of which above eight hundred were made upon the sub-currents.

The stations in the harbor, as well as those in the Hudson River, were designed to furnish us with additional data relative to a class of remarkable counter-currents discovered the previous season. The former observations had established the fact, that along the main channel the currents of the lowest water stratum maintain velocities and directions quite at variance with those near the surface. It however remained to be proved whether the phenomena observed were continuous from station to station or mere local conditions; and if their continuity could be shown, the exact limits of their domain were to be ascertained. The information now in our hands affords, I am convinced, a full and faithful exhibit of these points.

The three stations in the East River lie in positions which the previous work had shown to be important, as embracing the terminus of the Hell Gate interference current. At one of these stations, which lies in the deep basin westward of the point of Blackwell's Island, some curious conditions of the sub-currents manifested themselves. Here the axis of the ebb (westerly) drift was observed to lie about twenty feet below the surface throughout the entire duration of this current; in other words, the current is stronger at this depth than at any other point above or below. There are resemblances between this phenomenon and those already referred to as appearing in the main channel of the harbor, but I am doubtful whether we can class them together. In the discussion of our results, we propose to group the currents of the upper harbor according to tidal hours obtained from the self-registering gauge at Governor's Island, and those of the lower harbor and its approaches, according to tidal observations made simultaneously by some of our own party at Sandy Hook.

The closing work upon the physical survey of New-York harbor, which we have briefly described, did not occupy us during the entire season, and there proved to be ample time for the other operations directed by your instructions, viz., inquiries into the physical conditions of the bays and inlets along the south shore of Long Island.

Fur

Glancing at a chart of our coast, one may see on the south shore of the island of Nantucket a series of small ponds separated from the sea by narrow reaches of sand. On Martha's Vineyard the same features may be observed along the outer shore, except that here the larger basins or lagoons have occasional outlets through the strips of sand beach. ther to the westward, upon the coast of Long Island, appear similar basins, so extensive as almost to form inland seas with outlets of considerable depth, through which vessels may pass. Here are fully established the forms which may be distinguished as the leading characteristics of the Atlantic coast to the southward, and of the entire Gulf shore. From the past history of the sandy portion of our sea-coast, it appears that the outlets to which we have referred are never permanent, but continually shift their positions, either by gradual encroachments and recessions of the sand reaches, or by suddenly closing up at one point and breaking away at another. The design of our study was to ascertain, if possible, the causes which maintain these extended sandy reaches, and the agencies which create the channels through them. In this undertaking, the line of stations, to which reference was made in the former part of this report, extending along the coast, at intervals of from five to twelve miles from Coney Island to a point twenty-five miles east of Fire Island light, gave us all the requisite data for a complete knowledge of the shore currents and we added to these series of current observations at Fire Island, Crow Gut and Rockaway inlets, besides others from stations in the Great South Bay. By half-hourly records at gauges temporarily erected, the form of the tide-wave as it enters Fire Island Inlet was compared with that observed at Sandy Hook on the one hand, and that at the eastern extremity of the Great South Bay upon the other. At some of the outside stations we threw over sinking articles, hoping to find them again upon some portion of the beach, and thus be able to determine the direction of the movements on the bottom of the sea. The first class of articles we tried were balls made of cement, with corks enclosed, giving them what we

supposed to be the requisite specific gravity. On a former occasion these cement balls were used quite successfully along the shores of Sandy Hook, but we now found them to fail entirely on this coast, where the currents are more feeble. We subsequently had recourse to the large skimmer shell (mactra solidissima) which we collected from the shores, and marked with drills. Some of these, cast over in three fathoms water off Oak Beach, travelled eastward, and crossing Fire Island Inlet, were swept on shore four miles to the eastward of their place of deposit. In the performance of this journey they were occupied over two weeks, during a prevalence of easterly winds. Of three hundred shells cast into the sea, one hundred were recovered a much larger proportion than we could possibly have expected to find among the shifting sands and the miscel laneous stranded articles upon these beaches. The easterly preponderance in the movements of the currents along the bed of the sea, which the journey of these shells revealed, corresponds with the results from previous observations of the surface drifts.

As our inquiries proceeded, it became evident that the currents, powerful though they may be to scour channels and form the ocean bed, cannot alone effect the peculiar changes which are observed to take place in the beaches, but that the waves take a part, not insignificant, in these operations. In the shallow waters along these alluvial shores the waves, driving in from the ocean, acquire violent horizontal movements, and dash along the beach with a force in comparison with which the strongest currents are quite impotent. In order that we might the better understand and determine the precise action of the waves and the relation of their office to that of the current, we made a very careful examination of the conditions in miniature forms of bays and inlets, where the limited field of observations afforded us a comprehensive view of the natural activities at work. The results of this examination have already been laid before you, with such conclusions from them as were immediately obvious.

In the course of this and former reports I have referred repeatedly to the observations of currents at different depths, and it has occurred to me that some description of the apparatus in use should here be given, in order that the reliability of the results should be established.

DESCRIPTIONS OF APPARATUS.

For observations upon the surface currents we use a "tube-log," which is simply a tin cylinder four inches in diameter and six feet long. This tube is partially filled with water, so as to sink nearly its whole length and maintain an upright position; and a graduated line being attached, the observations are made as with an ordinary ship's log. We have found that a log of less draught than this is liable to be affected by the wind.

When we desire to obtain the mean motion of a stratum of greater depth, we use twenty-four feet feet tubes, and in some cases those as long as forty-eight feet.

If the velocity of the current in the lowest water stratum is desired, we take the following course: Two copper globes of equal dimensions are connected by wire rope of the smallest possible size compatible with the strength required. One of these globes, being filled with water, is allowed to sink the whole length of the connecting line, while the other being empty, or only partially loaded, swims at the surface of the sea. To the upper globe the log line is secured. The velocity with which the globes,

thus connected, will move, is a mean of the rates at which the upper and lower water strata are flowing; and if simultaneous observations are made with this apparatus and the surface log, before described, we are furnished with the means of obtaining by calculation the velocity of the lowest stratum. This method may be employed where the water is not so deep as to give to the connecting wire rope an extent of surface which, exposed to the current, may require consideration in the problem.

The instruments I have described thus far are similar to those which have been used in determining the discharge of canals in Europe. In the application of these to inquiries on a larger scale, I have found it necessary to make certain modifications of them to insure accuracy.

It not unfrequently occurs that the velocity of the surface drift is many times greater than that of the lower stratum, or holds altogether a reverse direction, so that the motion of the globes is quite at variance with that of the surface log. In a case like this, the graduated line secured to the globes is borne away by the surface current, and the observer is deceived. The full extent of this difficulty will be appreciated when it is considered that the line of which I have spoken is necessarily of considerable size, the strength of six men being sometimes required to draw in the globes. I propose to obviate this difficulty by the following arrangement: Within the upper globe, made of wood in this case, a reel is placed, upon which a small log-line, passing in at an aperture at the pole, is wound by a crank from without. The extremity of this log-line is secured to a third globe, which swims freely upon the surface of the sea. When making an observation, the log-line is wound up until the floating globes are brought together; then, at a signal, the reel is loosened; and now, if the surface and sub-currents differ in velocity, the free globe separates from the others, and the observer notices the number of divisions of the logline drawn out in thirty seconds. In this experiment the apparatus is in nowise connected with the vessel, but the observer follows in a boat until the trial is completed. Figure 1 (Sketch No. 40) shows the relative position of these globes during the course of the observations. In this figure A and B are the connected globes, while C is the free float. The weight of the globe B causes the swimming globe A to sink nearly to its pole, and the free float C is loaded so as to sink about the same distance. The graduated line, which measures the separation of the floating globes, may be seen, one end fastened to a ring upon the free float C, the other passing in at the pole of the globe A. Figure 2 is an enlarged representation of the globe A, opened so as to show the reel within. The water, which enters the globe freely, acts as a check, preventing this reel from acquiring an undue momentum with any sudden jerk of the line caused by the waves. In this figure may be seen the position of the crank by which the reel is wound; this crank is, of course, removed after the floating globes are drawn together. Upon the outside of the globe containing the reel every ten degrees are marked, that the observer may note in his record the amount submerged. In the reduction of these observations the extent of the wetted surfaces of the two connected globes must be considered, since, in the case of a difference of velocity between the upper and lower strata, the effective areas of the surfaces exposed to the two streams enter into the problem. In all positions the effective surface which a globe wholly immersed presents to the current is a great circle. The velocity attained by the connected globes is a simple mean of the