tending from Culebra and emptying into the Chagres on the west side of the canal below Gamboa. A survey party has been at work locating a diversion channel for the Obispo River, and other streams on the east side of the canal, which will extend from Culebra and will run approximately parallel to the canal, discharging its waters into the Chagres near Gamboa.'

CHANGES AT PACIFIC END.-The changes at the Pacific end are the most important made since the adoption of the original plans. The Chairman's report to the Secretary of War states:

"The adopted plan for the building of the Panama Canal contemplated the formation of a lake on the Pacific side by the construction of three earth dams (Sosa-Corozal, Sosa-San Juan and Corozal-Diablo), the differences of level between the lake and the Pacific being overcome by a flight of two locks projecting into the Pacific on the west side of Sosa Hill."

"The Board of Consulting Engineers that accompanied you to the Isthmus in April last, inspected the sites of the dams, as finally located, with a view to outlining a detailed description of the preparation of the foundations for the structures to be erected, reporting thereon as follows:

"The borings showed so-called mud in the swampy portion, having a depth of 8 or 10 feet. This material is firmer than we had expected, and at the time of our visit, about two days after it had been flooded by spring tides, it could be walked on in most places......For the dam construction .....we do not think it will be necessary to remove the soft material at any of these places (Sosa-Corozal, Sosa-San Juan). The very softest material will either be displaced or consolidated by the material disposed on it. We do not think the amount displaced beneath the impervious portion of the embankment will be great, and if compressed in place the material will be impervious."

As regards the character of the material that should compose the body of the structures, the same Board ported:-

or more.

"For the Sosa-Corozal and La Boca dams the bottom width of the impervious portion should not be less than would result from a top width of 80 to 100 feet at 15 to 20 feet above the water level in the lake, with slopes not steeper than 1 on 4 to 1 on 5. The resulting width at sea level would be 640 feet, The comparatively small cross-section suggested for the impervious portion of each of the Pacific dams is permissible only if it is reinforced by wide and heavy rock embankments on both sides...... On account of the nature of the bottom on which this rock fill will be deposited for the dams on the Pacific side, it should extend a long distance outside of the impervious portion in accordance with the principle adopted for the La Boca dam by the minority of the Board of Consulting Engineers in 1905-1906, of spreading the base on soft bottoms so that the change in weight imposed on the foundations from the toe towards the crown will be gradual. This protection at the SosaCorozal and La Boca dams may be given great width with economy since they will afford convenient dumps for the Culebra material."

re

To construct the dams in accordance with these views trestles were built along the toes of the Sosa-Corozal dam, from which to dump material from the Culebra cut. The trestles failed, after the dumping of material from them began, and the material overlying the rock moved laterally carrying the superimposed mass with it, the dumps flattening out until the side slopes were about 1 on 12 and even less. In places, this lateral motion continued for two weeks after dumping had stopped. The ground on either side of, and at some distance from the dump was forced up forming mounds of mud, the crests of which gradually approached the level of the top of the dump proper. After an equilibrium was established between the dump and the adjacent mounds, the hump or wave would again move out when the track was shifted towards it, accompanied by a sudden vertical

THE

SUMMER

HOUSE.

In front of the Market, Empire, Canal Zone.

C. D. KINKEAD, Proprietor.

SALOON AND TAILORING ESTABLISHMENT.

settling of the track of six to ten feet when loaded trains were applied. It was thought that if the trestle was lower, better results might be obtained, but with the west trestle only eight feet above the surface of the ground and the piles driven to rock, similar failures followed.

When the difficulties on the east trestle developed, an investigation by borings and test pits was made to determine the character of the material overlying the rock. It was found to be for the greater part of an unctuous blue clay without grit, possessing but little supporting power, instead of stiff clay as indicated on existing profiles.

The investigations clearly demonstrate that the construction of dams which would remain in place after the lake is filled can be accomplished only by the removal of all material overlying the rock, and after preparation of the rock surface, by building the dams of selected material. The depth to which the material would have to be removed is from 10 to 70 feet, and suitable material for dam construction would have to be transported to the site. Investigations of the foundations of the Sosa-San Juan dam show that the material is the same; the difficulties to be encountered will be greater because of the river.

It is estimated that the cost of building the two dams following the plan that the investigations and work done indicate as necessary will be about $11,573,800. The estimate submitted by the minority of the Board of Consulting Engineers for these dams is $4,314,000, a difference of $7,259,800. As this difference to secure stability

is excessive, an examination of the canal route from Pedro Miguel to the Pacific was undertaken to ascertain if more suitable places for the locks and dams could be found. In making these examinations care was taken to secure samples of the materials to be encountered at various depths, and cores procured of the rock. A careful study was made of all data obtained and four possible projects present themselves for constructing that portion of the canal extending from the south end of Culebra cut to deep water in the Pacific Ocean.

First Project. This is the present project and consists of one lock at Pedro Miguel and two at La Boca, with dams of suitable material on rock. The estimates submitted by the Board of Consulting Engineers are corrected to conform to methods, quantities and unit prices that additional information and experience show will give more nearly the actual cost of the work. The project provides a 500 foot channel from Pedro Miguel to Miraflores, a 1,000 foot channel through Sosa Lake to La Boca, and a 500 foot channel from La Boca locks to deep water.

Second Project.--Two locks at Pedro Miguel and one at Miraflores, with a 500 foot channel through the lake from Pedro Miguel to Miraflores, and a 500 channel from Miraflores locks to deep water in the Pacific Ocean.

Third Project.-One lock at Pedro Miguel, and two at Miraflores with a 500 foot channel through the lake from Pedro Miguel to Miraflores, and a 500 foot channel from Miraflores locks to deep water in the Pacific Ocean In this and Project No. 2, provision is made for diverting the waters of the Rio Grande and the Corundu and the Cocoli Rivers, as well as for a channel of uniform width to the sea, having no connection with any other tidal basin; as a consequence the currents in the channel will result only from the tidal flow and will not attain a velocity as great as one foot a second, which will not interfere with navigation.

Fourth Project.-One lock at Pedro Miguel, one at Miraflores, and one at La Boca, with 500 foot channels between locks and to deep water in the Pacific.

In each case locks and dams are on rock foundations. Assuming that inverts are used, the total cost of the various projects are as follows:

From the foregoing it is seen that Project No. 3, one lock at Pedro Miguel and two at Miraflores is the most economical. It has the advantage over the present project in that the dams of lower height, less length and resting on rock comparatively near the surface can be more easily constructed, and completed at an earlier date. It is to be preferred to Project No. 1 by reason of the fact that the location of the locks secures them against all possibility of distant bombardment and affords them greater security against gunboat or torpedo boat attack.

The Commission unanimously recommends the adoption of Project No. 3, and its substitution for the existing project."

The President approved Project No. 3 on December 20, 1907.

The two new locks at Miraflores will be built diagonally across the Rio Grande valley, connecting on the east with the hills at Miraflores and on the west with Cocoli Hill by short dams founded on rock. Recent investigations have shown that there exists a suitable foundation at the new location for the locks and dams contemplated. The locks lie directly across the valley and almost eliminate the question of dams, the upper end of the locks being so close to the Miraflores hills and the lower end so close to Cocoli Hill that the gaps can be closed by very short dams founded on rock. Under this plan there will be vir