This table is not accurate, but it is fairly reliable. Under the system of annual checking up, certain shoaling taking place between 26 and 30 foot depths might not show in the calculations for a 26-foot channel but show for a 30-foot channel.

The following are the amounts necessary to be dredged at the periods stated to produce the present project channel with depths of 21, 26, 30 feet.

It will be noticed the annual shoaling is materially higher when the 30-foot section is included than when the 26-foot section is considered. The 21-foot section is also included and this increases the amounts.

4. Objects of the survey.--In connection with other data already available, efforts were directed toward obtaining the following:

(a) Hydrographic survey with the condition of existing works of improvement.

(b) Discharge observations with current meters.

(c) Observations on the movement of silt, salt and fresh water, direction and strength of currents, etc.

Through these data study is made in an effort to arrive at a new project of improvement which will not require as much maintenance as the present one, and will arrive at full project dimensions which have not been obtained to date in either the 26 or 30 foot portions of the channel. Much of the detailed data and the methods of collection and computation are omitted from this report, as they would make it unduly long for anyone trying to get a comprehensive view of the situation. It was originally proposed to place this matter in an appendix so that the correctness of the methods could be checked, and also with a view to a permanent record which might assist others having similar work, but due to press of time this appendix was not prepared to go with the report. At the risk of sounding pedantic, the theory on which conclusions are based is given in many cases. This is only done so there will be no doubt of the line of reasoning used.

5. Drawings. Twenty-eight maps in triplicate, D. S. H. 135/1-28, inclusive, are inclosed."

The first sheet of the drawings, D. S. H. 135/1, is an index map of the following 12 drawings, D. S. II. 135/2-13, covering the general survey, and it shows in addition the general layout of the various

* Only 2 maps of this set printed.

channels, the location of the tide gages, and the cross sections, numbered from 2 to 22, where discharge observations were made. The location of the discharge sections at other points is indicated in the text when referred to. Drawings D. S. H. 135/14-15 give the general layout of the river section of the harbor on a larger scale but extending only from Kings Island to the end of the jetties at the mouth of the river. They show the stations which were put on for purposes of reference, and which are located 1,000 feet apart, starting at the foot of Kings Island with station 100 and running to station No. 206 at the outer end of the jetties. The named sections of the harbor (Upper Flats, Tybee Knoll, etc.) are all marked on these drawings, except Tybee Roads and Tybee Bar, which are off the limits and which show on the index map. Drawings D. S. H. 135/16-28 are charts giving various information in connection with the survey.

The following is a list of the index subnumbers and subtitles of all the drawings:

Sub-
No.

1. Index map.

2. Drakies Cut to sugar refinery.

Subtitle

3. Kings Island Channel and Head Back River.

4. Marsh Island Channel and city front.

5. City front, Wrecks Channel, and mouth Back River.

6. Obstructions, the Bight, head of South Channel, and St. Augustine Creek.

7. Upper and lower flats and South Channel.

8. Lower flats, Long Island Crossing, and South Channel.

9. Long Island Crossing, Horseshoe Shoal, and South Channel.

10. Quarantine Channel, Tybee Knoll, and mouth South Channel.

11. Tybee Knoll.

12. Tybee Roads.

13. Tybee Bar.

14. Upper half of river section.

15. Lower half of river section.

16. Surface slopes.

17. Location of jetties, Tybee Bar.

18. Vertical velocity curves at cross sections 8, 10, and 12.

19. Vertical velocity curves at cross sections 12A and 16.

20. Observed and mean velocities at cross sections Nos. 10 and 12.

21. Observed and mean velocities at cross sections Nos. 12A and 12B.

22. Observed and mean velocities at cross section No. 16.

23. Observed and mean velocities at cross section No. 17. 24. Analysis of water samples.

25. Graphical description of channels.

26. Graphical description of channels.

27. Effect of contraction in width on cross sectional area.

28. Details of proposed dams in cross tides and South Channel.

6. Tidal data.-Much of the tidal data secured were needed only for the discharge observations and are not given here. However, some of them have a more direct value in considering the improvement of the river and are given herein. Fifteen tide stations were established, with self-registering gauges. The data obtained are shown in a condensed form in the following table:

7. Discharge observations.-Unusual difficulty was experienced in making discharge observations and reducing them to a usable form, and no precedent could be found which was altogether satisfactory. One of the greatest difficulties was the irregularity of the flow for a particular tide. Of course, with the rising and falling tide the areas of cross section and velocities were constantly changing, and we took a number of observations to see if there was not some method of taking a few observations to give conditions throughout a cross section at a particular time, but we discovered no method. Even in a vertical filament we had to take a number of observations between the top and bottom. This resulted in out using a number of current meters at the same time for a particular observation and taking observations short distances apart in each vertical, of which there were seven in the wider cross sections of the river.

After a particular flood and ebb tide were measured we had difficulty in using the results. The fresh-water discharge of the Savannah River varies at its mouth from a minimum of about 3,000 cubic feet per second to a maximum of over 300,000, with a mean of 11,850. While there is a tidal range from mean low water to mean high water of 6.9 feet at the end of the jetties at the mouth of the river, this mean tide seldom occurs. The morning and the afternoon tides, according to the tide tables, should differ by as much as 12 feet in range on one day, while on another day they are the same. There is a large variation in the tides at different times in the month and this variation differs with the different months. Sometimes the tidal range is the same in two tides, but the level of the two highwaters (and hence low waters) is different, which affects the discharge. Combine now the influence of the barometer and the wind and we have new combinations. As the fresh-water flow changes, the matter is again changed, and as we move upstream the tidal influence becomes less and the influence of fresh-water discharge greater, until in front of the city the flow is sometimes not reversed during flood tide with a freshet in the river. The proportion of the flow that goes in different channels is also varied by winds.

But what is desired is the effect on the channel, and this is determined by the average conditions. All observations were therefore reduced to mean or average conditions as well as it could be done and the conclusions from these figures should be sensibly correct. A concave bank which is safe with normal flow, might wash with a large flow and fill with a small flow, and it might be necessary to guard against one or the other, but it would be almost impossible to guard against all such contingencies by theoretical discussion, nor is it considered practically important to do so.

8. Propagation of tidal flow.-Due to the various channels, the tide does not simply flow in, stop, and flow out, as in a single channel. The following is a statement of the main features of the normal flow and ebb: After low water in the ocean the rising flood begins to drive the water into Savannah Harbor by way of St. Augustine Creek, South Channel, and between the jetties in the North Channel. At the beginning of the flood, part of that entering between the jetties is diverted to the north of Long Island training wall, but this is later reversed so that the flow is from beyond the Long Island training wall into North Channel. This is because wing Dam 31 allows little flow across it when the tide is low. Following the flow up North Channel, we find that at Fields Cut, during the first quarter of the flood period, the water flows from North Channel into Fields Cut. During the next half of the flood period the flow is from Fields Cut into North Channel. During the remainder of the flood period, the direction of flow is again from North Channel into Fields Cut. The result is that the total flood volume passed during flood tide is not greatly affected by Fields Cut.

The flow coming up South Channel is met by the flow from St. Augustine Creek. The situation at the junction here is rather curious, in that the flood coming up St. Augustine Creek starts at the junction about 30 minutes earlier than it does coming up South Channel, so that during the first 30 minutes the water flows in from St. Augustine Creek and out through South Channel toward the sea. The flood in St. Augustine Creek, in addition to starting before the flood in South Channel, also keeps up about 40 minutes longer, so that at the end of the flood tide in St. Augustine Creek, water again comes up St. Augustine Creek and runs down South Channel. At the head of South Channel high water occurs about the same time, coming by the different routes, so that the two flood tides meet here but separate again, one passing up Back River and one up Front River. That passing up Back River moves more rapidly than that passing up Front River, so that high water and low water opposite Cross Tides both occur earlier in Back River than they do in Front River. The result is that the flood volume coming up Front River during the last 25 minutes is decreased above Cross Tides by the amount of water going out through Cross Tides and down Back River, and the last of the ebb flow down Front River is. increased by the first 14 minutes flood flow coming up Back River and through Cross Tides Channel.

Above Cross Tides the main river and the upper extension of Back River are sensibly single channels as far as tide flow is concerned, as the connection between them is small.

9. Rate of propagation of high and low water.The following gives the rate of propagation of high and low water for various sections of the river, as computed from our data:

10. Proportion of fresh-water flow for different channels.-The total fresh-water flow of the river based on discharge data was supposed to flow through the various channels as follows. It was assumed that the increase in each case in ebb over flood flow was fresh water. This is not necessarily true in each channel, but the total amount of fresh water in parallel channels is equal to the total difference between ebb and flood flow.

11. Preliminary discharge volumes.-The following table gives the volumes of the total ebb and flood tides at the various cross sections, in cubic feet, after reductions to mean conditions of tide and fresh. water flow but before final checking. The location of the numbered cross sections are shown in drawing D. S. H. 135/1.

1 Refers to water flowing north.

Refers to water flowing south.

The Tybee Sound discharge is approximate and is computed from known discharges, areas, and tidal ranges. It is the discharge across a straight line drawn between the shore ends of the two breakwaters discussed in par. 16 and shown on D. S. H. 135/17.