tion of the additional lock system considerably ahead of the time when it would be required for commercial needs.

6. Early in the studies conducted under authority of Public Resolution No. 85, the matter of special protection for the locks was given careful study and consideration, and the definite conclusion was reached that it would be a serious mistake to construct them immediately adjacent to the present locks. Since reaching this conclusion, efforts have been directed toward the development of a project referred to hereinafter as the "bypass project," in which the locks would be located at a safe distance from the present locks and approach channels would be provided to connect the new locks with the present canal channel, thus securing many of the advantages of two canals.

NEED FOR ADDITIONAL LOCKS FOR COMMERCIAL TRAFFIC

7. The need for additional lock capacity may depend on commercial traffic or on defense considerations. In the 1931 report, it was shown that the maximum daily capacity of the present locks during the 3 or 4 months' biennial overhaul period is 27 lockages in 24 hours, 26 of which would be available for commercial vessels, and that the capacity during the overhaul period is a limitation which determines the annual capacity. However, if this maximum daily capacity is used for determining the average capacity of the canal during the period of the overhaul, there would be excessive delays to traffic since experience shows that when the daily average reaches 27 lockages per day, there would be many successive days during the overhaul period in which the traffic will be greater than the maximum daily capacity of the canal.

8. The 1931 report suggested that "the economic procedure will be to take care of 'peak days' by delaying transits and compensating rebates." This is one solution, but further study of this matter indicates that the attainment of maximum economy will probably not be the most desirable procedure. Such a procedure would add so much uncertainty to ship schedules and operation and be so irritating to shipping in general that it is undoubtedly better to make some compromise provision for taking care of successive peak days during overhaul periods so as to give a reasonably adequate service to shipping. In accordance with this idea, it is believed to be safer to assume that the present average daily capacity of the Canal during the 3 or 4 months' overhaul period is 24 lockages per day throughout the entire period (of which 23 lockages would be available for commercial ships), or that the maximum daily capacity is about 12 percent above the average daily capacity.

9. Excluding transits of the fleet, the highest average daily number of lockages for any month up to date is 18.32, reached in the fiscal year 1929. During that month (December 1928), the average number of vessels per lockage was 1.18. Assuming that the average might have been increased to 1.48, as suggested by the Interoceanic Canal Board in its review of the 1931 report, the number of lockages could have been reduced to a daily average of 14.6 per day. Thus, with an average capacity of 24 lockages per day, the average daily traffic in the maximum month of record was about 60 percent of capacity.

10. On a tonnage basis, the capacity of the Canal depends on the average tonnage per lockage. For the fiscal year 1938, a normal year, the average tonnage per vessel was 4,035 tons, new measurement basis. Applying the factor of 1.48 vessels per lockage gives 5,972 tons per lockage. Thus the tonnage capacity of the Canal for commercial traffic, based on capacity during the overhaul season of 23 lockages daily, is 137,000 tons per day, or 50,000,000 tons per year.

11. To predict when the traffic will reach the capacity of the Canal is a difficult problem. It requires a broad knowledge of all the factors that affect the growth of world commerce and shipping, and no prediction however carefully made, can be accepted with complete assurance. Fortunately, we have a recent prediction of future traffic which we consider to be the best obtainable with the information now available.

12. In connection with the investigation of Panama Canal tolls, rates, and measurement rules, made under authority of Congress in 1936, a study was made of the future growth of traffic at the Panama Canal. The report on this study is printed in Senate Document No. 23, Seventy-fifth Congress, first session. This study, however, predicts only the average growth in traffic and not the traffic that may be expected in any particular year in the future. To obtain this from the traffic curves used, it is necessary to apply a factor for secular variation. During the year of maximum traffic, 1929, the total tonnage

on the new measurement basis was 27,600,000 tons. This was 4,600,000 tons or 20 percent in excess of the traffic that might have been expected in that year based upon the average normal growth of Canal traffic for the 20 years covered by the study. In other words, our past experience had indicated that the maximum annual tonnage in any 1 year may be about 20 percent greater than the average growth would indicate. Obviously, in predicting the annual tonnage for any particular year in the future the safe thing to do is to assume that the year will be at the peak of a "boom" cycle. Thus the tonnage to be used in connection with the curve of growth to determine when the Canal capacity will be reached is the annual tonnage capacity of the Canal, 50,000,000 tons, reduced by application of the above percentage, or 41,670,000 tons. This tonnage applied to the curve of future traffic shows that the capacity of the existing locks may be reached by 1961. Additional locks would be needed, therefore, by that date.

NEED FOR ADDITIONAL LOCKS FOR DEFENSE

13. The Canal must be kept open for transit of the fleet from ocean to ocean, and the locks, therefore, must be given the maximum possible physical protection from vital damage. It is, of course, apparent to the casual observer that structures such as the locks are subject to attack by sabotage or by air raids and that, if an additional system of locks were constructed, the chance of vital damage by such attacks resulting in serious closure of the Canal to traffic will be reduced. If, further, the additional locks were constructed in locations removed from the present locks, the chance of such vital damage would be further reduced. Detailed studies show that a great deal can be done in the design, construction, and operation of locks to give them protection from vital damage by air raids. Furthermore, closing new locks and the bypass channels to all traffic except the Navy, would give the highest practicable degree of assurance that no attack on the locks could close the Canal for more than a very short period of time.

14. Thus the provisions of the additional locks and bypass channels will increase the security of the Canal, and these additional facilities will be needed in any event for commercial purposes within a relatively short period of time.

ENGINEERING INVESTIGATIONS

15. With the funds appropriated under authority of Public Resolution No. 85, an initial organization and plans for carrying on the work have been developed and office studies have been made. In the field, a topographic map on a scale of 1 to 1,000 has been completed. This map covers the areas within which lie all the desirable locations for the proposed locks and bypass channels. Borings under the supervision of a competent geologist have been made at all lock sites to determine suitable foundations.

16. Investigation in the Gatun area: Surface observations: Topographically the Gatun area is a ruggedly dissected plain of low relief, 100 to 200 feet average, which breaks off somewhat abruptly to a low coastal marsh bordering Cristobal Bay to the north. Rock outcrops along highway, railroad, and Canal banks indicate the underlying rock consist of a series of sandstone and claylike beds dipping gently to seaward (northward).

Subsurface studies: Two lock sites were drilled at Gatun, one roughly parallel to and about 3,000 feet removed to the east of the present locks, line A-7, and the other, line A-1-A, parallel, to and about 500 feet east of the present lock. Nineteen holes were drilled at the proposed lock position along line A-7 and nine holes along line A-1-A. Roughly, 400 to 500 feet of rock strata underlying the immediate lock area were penetrated. Conclusions reached during surface reconnaissance were closely borne out. By far the largest portion of the rock strata is medium hard, very fine-grained sandstone, with interbedded claylike beds of volcanic ash, tuff, and pumice fragments, and several thin beds of coarse, gritty sandstone conglomerates. Due to the dip of the beds, the northern two-thirds of the lock would rest on sandstone, the southern third on alternating beds of volcanic ash, tuff, sandstone, and conglomerate. Only a very thin zone of weathered rock is present over the area, averaging perhaps under 6 feet. Practically all of the rock below this weathered zone is firm to medium hard, and sound. The bedrock is mantled with yellow and red silty and sandy clays of variable depths, averaging perhaps 20 feet more or less.

17. Investigations in the Pedro Miguel area: Surface observations: The area in which it is proposed to found the third set of locks at Pedro Miguel lies

immediately west of the present locks. The proposed lock site lies roughly parallel to, and about one-quarter of a mile removed from, the present locks. Topographically and geologically this area is typical of the region along the Pacific side of the Canal Zone, being characterized by numerous steep-sided conical-shaped hills of igneous, usually basaltic rock, rising to elevations of 100 to 700 feet above the surrounding country. These knobs usually occur in "nests" of several, with flat-floored river valleys intervening between the groups of hills. These flat valleys are usually underlain by sedimentary strata of interbedded shales, sandstone, conglomerates, volcanic ashes, and consolidated volcanic mud flows. The proposed lock site lies in the flat-floored valley of the Rio Grande, which here is bordered by high basaltic cones on the south, west, and north.

Subsurface studies: Sufficient core drilling has been done to indicate that a set of locks constructed here would be founded on substantially hard basalt, sandstone, and shale. Both ends of the lock would rest on basalt and the central portion, where it crossed the flat Rio Grande Valley would rest on very hard sandstone and some medium-hard shale. The rock is weathered and shattered to variable depths, but at all points explored with the drill, the top of firm, sound rock is well above the foundation level of the proposed lock. The overburden in the area consists of clay, silty sand, and small rock fragments with a few scattered boulders. Foundation rock has been explored to depths of over 100 feet below the proposed lock foundation level.

18. Investigations in the Miraflores area: Surface observations: The proposed third locks at Miraflores lie roughly parallel to, and about one-fourth mile removed to the west of, the present Miraflores locks. The site selected lies in a nest of volcanic rock cones and low hills. By far the largest part of the lock would be founded on hard, basaltic, igneous rock. The ends of the lock extend a short distance beyond the basalt hills the north end out into the Cocoli River Valley, and the south end out into a low swamp area.

Subsurface studies: The drilling has indicated that the locks will be founded on sound, igneous rock.

19. Seismic studies: In order to facilitate subsurface studies and assist the geologist in arranging an efficient drilling program, a geophysical specialist was obtained from the office of the Chief of Engineers, United States Army. Seismic studies have been coordinated with core drilling records in a practical and useful manner.

20. Sources of aggregate: Investigations indicate the presence of vast quantities of river gravel in the Chagres River Valley between the town of Gamboa and Madden Dam. All lock sites can be reached from there by water haul. Crushed rock, hard basalt, can be obtained at Gatun by quarrying on several of the islands in Gatun Lake near Gatun. The basalt hills at the Pacific locks sites are suitable sources of hard rock materials.

21. Cement materials: A study has been made of the limestone deposit along the Massambi River near Summit in anticipation of a request for a source of suitable materials for cement manufacture. Recently studies have been made of the vast deposits of corals in Cristobal Harbor and bordering the swamp areas around the harbor. Both deposits have been sampled and analyzed and have been declared suitable cement material. Recovery of the coastal coral deposit would be a dredging operation and probably far less expensive than quarrying on the Massambi.

22. Conclusions as to foundations: As a result of foundation investigations at the proposed lock sites, the following conclusions are reached:

(a) Locks constructed at the proposed sites at Gatun, Pedro Miguel, and Miraflores would be founded on sound rock.

(b) Foundation studies have been extended to include a sufficiently large area to permit minor shifting of the lock sites to obtain better approach channel alinement and to cover an area adjacent for a possible future fourth lock installation, without altering the conclusions reached in regard to foundation condition.

(c) No major rock profile irregularities or adverse foundation rock conditions, such as badly shattered fault zones, deeply weathered rock channels, overhangs, sharp folds, etc., have been encountered at the proposed sites.

(d) Adequate sources of concrete aggregate materials as well as most of the ingredients necessary for the manufacture of cement for the new locks are obtainable within the confines of the Canal Zone.

THE BYPASS PROJECT

23. Locations of the principal features of the project now proposed are shown on the accompanying maps. At Gatun, a set of locks will be built with three single chambers in flight, and approach channels will be excavated connecting the locks with Gatun Lake and with the existing Canal channel near the head of Limon Bay. The selection of the most desirable location in the vicinity of Gatun has been simple. Good foundations are obtainable at numerous locations to the eastward of the present locks. Suitable foundations are not available on the westward side of the locks, at least in the area between the locks and the Gatun spillway, and all locations beyond the spillway are impracticable on account of the necessity of providing either a new channel to the sea or one crossing the Chagres River, the fluctuating stages of which would not permit the maintenance at reasonable cost nor the satisfactory navigation of any such channel. Consequently, it remained to select the best site on the east side of the present locks. The site selected is about one-half mile from the old locks. It is at a desirable distance from them, and is so located as to avoid the relocation of existing structures as much as possible. The locks themselves are located as far to the north in the bypass channel as foundation conditions permit, so as to reduce the cost of excavating the channel by including as much of it as practicable in the high-level reach.

24. On the Pacific side the topographical features obviously make it impracticable to construct another set of locks on the east side of the Canal at any considerable distance from the present locks. In view of the separation of the total present lock lift into two parts at Pedro Miguel, one flight, and Miraflores, two flights, respectively, with Miraflores Lake between, it seems very desirable to preserve the flexibility of that arrangement in the new locks. To do so requires that the corresponding separation be made in the new lock lifts and that the channels between the two enter, or at least have access to, the present Miraflores Lake. This arrangement permits a vessel in the course of its transit of the Pacific locks to use any flight of lower locks with any of the upper locks. The advantages in the event of the closure of one or more locks for repair, overhaul, or by accident of any kind, are apparent. It has been found practicable to preserve these advantages in the locations now proposed and satisfactory foundations exist at the separate locations shown on the map. 25. At Pedro Miguel, a single-chamber lock will be located about one-quarter of a mile to the westward (as referred to the east and west side of the Canal, actually nearly due south), of the present Pedro Miguel locks. The north approach channel will pass westward of Cerro Paraiso to connect with the existing channel about two-thirds of a mile south of Gold Hill and about 2 miles north of the present Pedro Miguel locks. The southern approach channel will connect with Miraflores Lake about one-third of a mile below the present Pedro Miguel locks.

26. The new Miraflores locks, on the same side of the Canal, will consist of two single chambers in flight, with the northern end almost opposite the southern end of the present Miraflores locks and about one-quarter of a mile away. The upper approach channel will connect with Miraflores Lake at a point opposite the upper end of the approach wall of the present locks. The lower approach channel will connect with the main Canal channel opposite Balboa Harbor.

27. Detailed locations will be made when the detailed plans are drawn. The flexibility of operation secured by having the new locks on the Pacific side divided into two sets but with approach channels into Miraflores Lake, and the added protection secured by such dispersion, make the divided system more desirable than a single set of locks with three chambers in flight.

28. All locks will be specially designed and constructed to resist air attacks and sabotage. The lock chambers will be of such dimensions as will provide for such future growth in sizes of vessels as can be reasonably foreseen. Recent studies of the probable needs of commercial and naval traffic during the life of the new structures indicate that the lock chambers should be substantially larger than formerly proposed in the 1931 report. The estimates herewith provide for chambers with usable dimensions 1,200 feet long and 135 feet wide with 45 feet navigable depth, and with approach channel having a minimum width of 300 feet and depth of 40 feet at low water. The exact dimensions will be fixed when detailed plans are drawn. Provision will be made at Gatun and at Miraflores for a highway underpass or bascule bridge. At Gatun the Panama

Railroad and the electric transmission line will be relocated as necessary to avoid interference with the new channels. No increase is needed in the existing supply of water for operation of the proposed locks.

ESTIMATE OF COST

29. The following is an estimate of cost of the works proposed:

30. In order to compare the above estimate of cost with the cost of a project in which the new locks are placed immediately alongside of the existing locks as proposed in the 1931 report, it is necessary to add to the estimate in the 1931 report the additional costs of providing for a lock width of 135 feet instead of 125 and navigable depth of 45 feet instead of 42% feet, and of providing the special protective features, which would have to be incurred in any event if the plan proposed in the 1931 report were adopted. Thus, the estimate in the 1931 report would be increased to $219,000,000, so that, the bypass project is estimated at a cost of $58,000,000 more than would be required for a project with the new locks alongside the present locks as proposed in the 1931 report.

31. A large amount of preliminary work, including the relocation of railroads and other structures, the provision of rail and highway access to the work, and the extension of housing, water, light, power, sewers, and sanitation to the areas that will have to be occupied by the construction workers, can be undertaken whenever funds are made available. In addition, excavation of approach channels could begin at once. $15,000,000 can be used for these works during the first year.

SITUATION AFTER COMPLETION OF PROJECT

32. If this project is authorized now it will be completed at a time when traffic will still be well within the capacity of the present locks during the overhaul period, and thus the new locks can be held for the exclusive use of the Navy, except possibly for occasional commercial vessels too large for the present locks. After a considerable period of years, the traffic may reach the capacity of the present locks in the overhaul season, and thereafter, during that season, it will be necessary to use the new locks for commercial traffic. This period will last an indeterminate number of years because, by using the new locks during the overhaul season, the average capacity of the old locks on an annual basis will jump from 24 lockages per day to about 41 lockages per day (85 percent of maximum capacity of 48 lockages per day) except for occasional excessive peaks. 33. If, in the future, the number of commercial vessels to use the Panama Canal, of size that will not pass through the existing locks, reaches an important proportion, another set of large-size chambers could be built alongside the new locks proposed in the bypass project. It is probable, however, that this contingency will not arise for many years. Prior to that time, the use of the new locks for occasional large commercial vessels and during the overhaul need not affect their safety for defense purposes, because during the period of any emergency, the overhaul can be postponed and the new locks closed entirely to commercial traffic. Furthermore, it is proposed to design the new locks so as to require overhaul only at long intervals of time and so as to reduce the length of the overhaul period very greatly below that now required for the existing locks. 34. Operating costs on completion of the project will be increased relatively very little since operation of the new locks can be handled by the present force for many years. Some addition to the force will be required for maintenance work, and as time goes on the maintenance will increase. When traffic reaches a point where the new locks must be used during the biennial overhaul season, it will be necessary to bring to the Isthmus for the overhaul work an additional number of mechanics to replace our own men who ordinarily assist in overhaul work but who will then be required for operating the new locks, thus increasing the cost of overhaul.

35. If the project costs $277,000,000, as estimated, it will be necessary to add all or a part of this amount, plus interest during construction, to the capital