This work is in the Middletown collection district.

During the fiscal year ended June 30, 1872, the duties on goods in warehouses was $12,661.94.

In 1836 the United States appropriated $20,000 for deepening the water on this bar. It is believed to have been spent in dredging, but there are no effects of such work now discernible, and the exact plane at which it was done is unknown to me.

Very respectfully,

G. K. WARREN,

Major of Engineers and Brevet Major-General. U. S. A. General A. A. HUMPHREYS,

Brigadier-General and Chief of Engineers, U. S. A.

HARTFORD, CONNECTICUT, December 16, 1872.

GENERAL: In accordance with instructions received, an examination has been made of the mouth of the Connecticut River and Saybrook Bar during the past season, to determine whether any improvement could be effected, either by temporary or permanent works, which would deepen the water in the channel and benefit navigation over the bar.

The field-work was commenced in the early part of May, and completed about the middle of August. A party was organized under Mr. Charles L. Burdett, principal assistant, consisting of two sub-assistants, viz, Mr. Norman W. Eayrs and Mr. Luther W. Burt, to read angles to the points of soundings, four boatmen, to row the boat from which soundings were taken, one leadsman, and one extra man in the boat to manage the signals; in all, nine persons.

This party worked with very little interruption during the time mentioned, and with little change in the party, the same men remaining through the survey. The triangulation shore surveys and levels were made by Mr Burdett, who also went in the boat when the party were sounding and recorded the depths and times.

Mr. Eayrs and Mr. Burt, in addition to reading angles to the soundings, made some of the details of the shore survey.

The signals in the boat were managed by Mr. M. T. Burnett, who also served as spare oarsinan when occasion required.

The soundings were nearly all taken by Mr. Wilmot Hoxton.

The boatmen were Oliver Denning, George A. Paine, George V, Wilson, and George W. Newton.

The current-experiments were mostly taken by Mr. Burnett under my direction. The shore surveys were generally made in weather unfavorable for sounding, the whole party serving in such capacity as occasion required. There were, nevertheless, many interruptions from the uncertainty of the weather on the bar. It would often commence blowing, or a fog would come up, after the party had made all their dispositions for sounding, with the observers generally on different sides of the river, and much time would be consumed in waiting for it to clear or to get the party together for shore work.

Method of making the surveys.

The surveys were made by running base-lines near the shore, with such angles as would make them of suitable length and follow sufficiently near the bank to measure effects. These lines were run along the shore of the sound from Guard House Point to Day's Hill, and on both sides of the river to the railroad-bridge at Lyme.

All the points were connected by triangulation, and such surveys made as would show the characteristics of the bank for some distance inland.

The soundings were made from a twenty-two feet row-boat, and were taken as nearly as practicable on ranges determined by setting up two poles in line on shore, or by taking some well-defined objects to supply their place when they were too far off to be seen. Two theodolites were placed upon points of the survey which would make the most favorable angies for the line to be run, and read at the dropping of a signal upon the boat, whose motions they followed. The soundings were taken at every fifth stroke of the oars, and the signal was raised at every fourth and dropped at every fifth sounding. At the instant of dropping, the angles were taken and recorded, with the exact time of the observations. The time and the sounding to the nearest tenth of a foot were also recorded in the boat.

The signal was a flag upon a pole, weighted at the bottom so that it would stand upright. A line was attached to near the middle of this pole, and carried through a pulley at the top of the boat's mast. The line was then carried aft under the thwarts, and was operated from the stern of the boat while the leadsman stood at the foot of

the mast, under the signal, in the forward part of the boat. This signal was easily operated and proved very convenient. At times, when one man was absent, Mr. Burdett steered the boat, gave the signals, and recorded the observations of soundings and time without difficulty. Owing to the strong currents at the ebb and flow of the tide, the boat could not at all times be held up to the lines of the ranges on shore, but the angles taken by the observers with the theodolites determined with great exactness the position of every fifth sounding, and, with the exception of some few characteristic intermediates, these are all that have been placed upon the map.

The observations that were made of the force and direction of the currents upon the part of the bar at the present time used as the steamboat channel, were taken with a carefully-rated current-meter for obtaining the velocities, and with a vane arranged upon a pole of sufficient length so that it could be set at different depths to indicate the direction of the current. All of the positions were located by angles from the shore, Tide gauges.

Tide-ganges were established at such places as were necessary to determine the required data for reducing the soundings to mean low water.

In order to record the rise and fall of the tides continuously at some place near the bar during the time occupied by the survey, a permanent self-registering gauge was established at the southeast corner of the depot of the Valley Railroad, at Saybrook Point.

This was the nearest point to the bar that could be occupied without a permanent structure to protect the gauge from the force of the waves. It is about a mile from the bar, and the tidal wave could pass into the mouth of the river directly toward it. Through the kindness of Mr. Walkley, the president of the railroad company, I was enabled to place the box of the gauge, containing the recording-apparatus and the clock, inside of the passenger-room of the station, where it was secure from the weather and all interference.

The building stands near to the water, which passes under the wharf in front of it. up to the foundation wall. The vertical tube containing the float was attached to this wall opposite to the box inside and about five feet horizontal distance from it. Not being able to procure a self-registering gange from the Government, I constructed one something upon the plan of Saxton's, but it was made to register to a scale upon profile paper instead of the plain paper ordinarily used. By this means the heights could be more easily read upon the sheet, and the height indicated by the pencil could be at any time compared with the reading of a fixed gauge. The average difference of range from high to low water between the self-registering and the fixed gange was .036 of a foot; the range of the self-registering gange being the greater by that

amount.

A permanent gauge was fixed to the southwest corner of the same building for checking and comparing the self-registering apparatus and determining its base. The height and time were marked off nearly every day upon the record-sheet.

A gauge was established outside upon the bar, at the point marked G upon the map. but it was much disturbed by the waves, and was once carried away. Its indications are not so reliable as those of the gauges inside, but it was used to compare the heights and times of the tides inside and outside of the mouth of the river. A tide gange was also established at Lynne bridge, to assist in reducing the soundings in the river to the plane of reference.

All of these gauges were connected with permanent bench-marks. It is to be regretted that previous surveys were not connected with permanent points so as to compare the deduced levels of low-water mark and connect the tidal observations.

Mean low water.

The mean low water, or plane of reference to which the soundings are reduced, is based upon the means of two consecutive levations or synodical months of 29, 53 days. commencing from high water on May 20 and ending on July 18.

The average high and low water, and the mean levels for each levation, including fifty-seven complete tides, as registered by the self-registering gange at Saybrook Point, were found to be as follows:

This gives the mean rise and fall at Saybrook Point 3.203 feet. A comparison of heights of the tides at Saybrook Point and at point marked G on the map, outside the light, and south of New Saybrook, gave the average high water at G .01 of a foot higher, and the average low water 23 of a foot lower, than at Saybrook Point.

This gives the mean level of the sound .135 of a foot lower than the mean level at Saybrook Point, or 99.910, and the mean rise and fall of the tides, 3.493 feet.

A comparison of the gauges at Saybrook Point and Lynne bridge, about two miles above, showed that the heights of the tides were essentially the same at these points. In reducing the soundings, the mean low water outside was taken for soundings on the bar, and the proper proportion higher in the river.

The most accurate and easily determined plane of reference is undoubtedly the mean level of the sea, as the tides at all seasons and times of the month rise and fall nearly equally above and below it, so that a limited series of observations determines it with great exactness, whereas the low tides vary with the relative position of the sun and moon, with their respective declinations and their distances from the earth.

Their quantities are so complicated by the inclination of the orbit of the moon to the ecliptic, the rotation of its nodes once in about 347 days, and the rotation of its line of apsides once in about nine years, that it would be exceedingly difficult, if not impossible, to deduce precise mean low water from a two months' series of observations. The practice, however, being common, to use this base upon charts, it has been thought best to make it the plane of reference, and at the same time give the mean level to compare with future surveys; the mean low water taken being sufficiently accurate for our purpose. The time of the rise and fall of the tides at Saybrook Point, taken from the averages of the two lunations before mentioned, is as follows:

It would seem from the usual law of waves entering an estuary, that the front side of the wave should be the steeper, and the time of rise of the tide should be less than that of the fall; but such does not appear to be the case at Saybrook.

The time of high water on the bar at the point G precedes the high water at Saybrook Point by about six minutes, and the time of low water on the bar precedes the low water at Saybrook Point by about three minutes.

This makes the time of rise 6 hours and 17 minutes, and the time of fall 6 hours and 8 minutes. At Saybrook Point, the greatest rise of tides above mean low water is 4.30 feet, and the greatest fall of tides below mean low water is .95 foot, making the greatest range 5.26 feet. At the bar, the greatest rise of tides above mean low water is 4.59 feet, and the greatest fall of tides below meau low water is .96 foot, making the greatest range 5.55 feet.

Previous surveys.

The earliest survey of Saybrook Bar, of which we have any record, was a very rough one, made by Mr. John Copp, in 1720. The map was sent to England, but a tracing of what remains of it is in the Conuecticut State library. It only shows the general shape of the mouth of the river; that Poverty Point extended out much farther than at present; that the main channel was to the southeast, and had a depth of two fathoms of water.

The next survey of which we have any knowledge was made by Captain Abner Parker, of Saybrook, in 1770. A chart was made in 1771 from his survey, a tracing of which is appended to this report. It shows the different channels existing at that time, with the lines of direction and ranges for entering the harbor. It is very valuable for reference, as it is the best representation attainable of the bar as it existed one hundred years ago. The original was found in a very dilapidated condition, in the possession of Hon. J. H. Trumbull, president of the Connecticut Historical Society, by whom it was kindly loaned to me. The parts have been put together and mounted for preservation, very little being found wanting. It is believed to be the only copy now in existence.

A survey of Saybrook Bar was made in 1836, by Julius W. Adams, now of Brooklyn, New York, under the direction of Captain William H. Swift, of the United States topographical engineers. The orders for the survey were issued from the Bureau of

Topographical Engineers, August 23, 1836, and the report of Captain Swift was received on the 9th of February, 1837.

I have not been able to find this report among the printed Government documents, but presume, from the map, that dredging alone was recommended.

The original estimate for this work was $54,380.

Surveys were made at the mouth of the Connecticut River by the United States Coast Survey parties, in 1835-36-'45-'49-'50, and 1851, the results of which are found upon the chart of mouth of Connecticut River, published in 1853. The hydrographic part was all done in 1849 and 1851.

There is a map in the Coast Survey Office of the same upon a larger scale, viz, robo a tracing of which was appended to my report upon surveys of the Connecticut River of January 11, 1868. This, however, does not show anything more than is delineated upon the printed chart, a copy of which is appended, except that the soundings show something more of the channel of deeper water south of Griswold's Point than is clearly defined in the present survey.

In July, 1868, some soundings were taken on the bar by direction of Colonel D. C. Houston, of the United States Engineers, but no regular survey was made.

The tides in Long Island Sound.

The nature of the tides and tidal currents in Long Island Sound may be stated briefly as follows:

The tidal wave from the ocean approaches the entrance to the sound from the southeast and flows into it past Fisher's and Plum Islands, mostly through the middle part, called the Race.

Another tidal wave enters the west end of the sound through New York Bay. These tides meet at some point between Sand's Point and Throg's Neck, according to the state of the tides.

The wave passing up the sound from the race is partly a wave of translation, with a progressive velocity depending upon the mean depth of the channel, and partly an actual flowing in and out of water through the race.

The crest of the wave of translation moves from the race to Sand's Point, a distance of 95 miles, in two hours. As the tide rises, the water pours through the race at the maximum rate of about five miles an hour. This current, opposite the mouth of the Connecticut River, is about two and a half miles an hour. As the crest of the wave passes, the currents set back again in the opposite direction, and with about the same velocities.

As the tide rises in the sound the currents set in toward the shore, and as it lowers they set outward from the shore, taking a direction slightly diagonal to the general direction of the central current.

The bar.

The bar at the mouth of the Connecticut appears to be largely composed of detritus brought down by the river in times of freshets, although some portion of it is washed from along the shore to the eastward by the action of the waves, and some is probably swept across the mouth of the river by the littoral tidal currents of the sound.

During the spring freshets the water is highly charged with silt, and a great quantity of coarser material is probably carried along with the current near the bottom, which all passes out upon the bar. Its character is there modified by the action of the waves, and by the tidal currents flowing in and out over the bar, when the water in the river again falls to its usual level, so that only the coarser parts remain; the finer particles of silt being swept away. That part of the bar which remains nearly permanent, or subject to gradual and progressive changes, is mainly composed of coarse sand and other heavy material, much more difficult to move than the ordinary river deposits, which are found in the lower portions of the river-bed.

The prevailing direction of the waves that roll in upon the bar and the shores to the east and west, being from the southeast, a large amount of material is swept along the beach by the wave-action toward Griswold's Point and Poverty Island, and is constantly adding to the spits that extend outward from the eastern shore, until their length becomes such that they are washed away by a freshet or a gale from another direction, when the material is carried by the river-current out upon the bar.

This action of the waves, coming from the southeast upon the beach, will explain why there are no such spits as have existed at Griswold's Point and Poverty Island upon the western side of the mouth of the river.

The form of the outer edge of the bar is probably modified by the strong littoral tidal currents, which flow back and forth in this part of the sound. The effect of the tidal current from the westward is more marked, however, thau that from the eastward, as it seems to limit the extension of the bar in a southwesterly direction by eroding its outer edge.

We thus have the following causes operating to create the bar and modify its form:

1. The suspended silt brought down by the river in freshets, which deposits upon the bar by the decrease of velocity in the stream.

2. The sand and coarse material which are swept along the bottom and roll out upon the bar until they meet the water of the sound.

3. The wave-action which carries material along the beach toward Griswold's Point, to be in time carried out upon the bar by the river-current.

4. The wave-action affecting the outer slope of the bar to modify its form.

5. The littoral tidal current flowing up and down the sound, and the current flowing in and out of the mouth of the river, which erode certain parts of the bar, and probably deposit material upon other parts, and otherwise shift and assort the material brought down by the river, washing away the lighter particles, and thereby modify the general form and character of the bar.

The action of the tides at Saybrook Bar.

At dead low water the current from the river spreads out over the whole bar. The greater volume of the current, however, follows the central channel.

When the rising tide commences to flow in from the southeast it first dams back the river-water and forms a distinct line of ripples along the outer edge of the bar. As the tide continues to rise, the stronger flow sets this ripple farther and farther on to the bar, turning back the current from the river and throwing it farther to the westward, until what still flows out passes through the west channel, or the one now mostly used.

From half-tide upward a strong current sets in along the shore toward Griswold's Point, and flows across the mouth of the river and partly into it, setting the fresh water back and raising the height of the water in the river.

Just before the highest water the current turns and sets strongly to the eastward along the shore west from the light-house, impinging strongly upon the current flowing out from the river, and forming a well-defined ripple extending southwest from the light. The great volume of water coming from the river that has flowed into it on the rising tide maintains its current southward until the tide has somewhat fallen, when the great volume of water sweeping down the sound overcomes the obstacle and pushes the river-current over to the eastward into and beyond the main central channel.

This action of the littoral currents is undoubtedly what forms the two side channels shown on the map. It will be observed that along the southwest edge of the bar the outer bank is, for a long distance, very steep, the 6, 9, and 12 foot contour lines being quite near together. This is probably caused by the strong wash of the current of the ebb tide flowing along the shore from the westward and striking upon the outer edge of the shoal.

The channels over the bar.

At the present time there are three distinct channels over the bar-one to the eastward along the shore south of Griswold's Point, which is the shoalest of the three, and never used; one to the southeast, which is the deepest of all, and but rarely used; and one to the southwest, which is now the steamboat channel; and their course, as shown on the map, is followed by nearly all vessels entering or leaving the river.

It will be observed that this channel, now almost exclusively used for passing over the bar, is not the deepest and widest nor what would appear from the chart at the present time to be the best. A direct line, entering the river upon a single straight course over the bar, as shown by the dotted line on the map, nearly in range with the light-honse and beacon, gives about one foot more water than the channel at present used. Vessels could now easily pass through this channel, by keeping just to the eastward of a range from the light-house to the Saybrook Congregational church.

When this old channel was previously used as the main entrance to the river, before the year 1853, it was usual to make an angle in entering, as shown by the sailing directions on the Coast Survey chart. This angle, the point for making which could not be readily ascertained in the night when the New York boats cross the bar, and the strong lateral tidal currents, were probably the causes which led to the abandonment of this channel for the shorter and safer distance in a direct line across the bar to the southwest, where the boats now run.

The southwest channel, which has been exclusively used for the past nineteen years, lies in the direction of the greatest amount of river traffic, and is the nearest course for the New York boats, two of which pass over the bar each night. It also permits the approach of vessels to within sight of known ranges on shore before reaching the shoalest part of the bar.

The changes in the bar.

An examination of Captain Parker's chart of 1771 shows that there existed at that time four principal channels over the bar. Three of them correspond very nearly in posi