same form and size as the canal, except the width on top, which is 7 feet.

In many cases, it has been necessary to make provisions in the estimates, (as may be seen by reference thereto,) for protecting the banks of the canal from the encroachment of the streams. Wherever stone could be procured with facility, and the foundation was such as would ensure the permanency of the work, the estimates have been made with the view of erecting permanent stone walls on the exterior of the banks. With a single exception, this is the case at all the points exposed to the abrasion of the streams from Olean to Mount-Morris, embracing two thirds of the whole distance requiring protection.

On the Genesee river, below Mount-Morris, docking has been estimated for, in place of stone wall, on account of the scarcity of stone, and the general prevalence of quicksands in the bed of the river. The first cost would be less than that of a stone wall, but it is believed the latter would be the cheaper in the end, inasmuch as the interest on the difference in the expense will not pay for the repairs of the former.

On canals located in valleys of much depth below the summit of the surrounding uplands, and at the foot of steep side hills, the embankinents and other works are often in jeopardy, and frequently sustain great injury from the sudden swelling of the streams in the time of freshets. The Rochester and Olean canal is not greatly exposed to these torrents; yet a portion of this route is so sitaated, and to avoid these difficulties, (so far as practicable,) I have availed myself of every opportunity presented by the locality, for passing the flood waters under the canal by culverts and aqueducts. Waste-weirs have also been estimated fully adequate to draw off the surplus accumulation of water in the canal.

The plans adopted for the mechanical work on the canal are generally of the same character as those on the Chenango canal.

The variation which occurs in the estimates of the several structures is partly owing to the difference in the cost of procuring the materials and partly to the quantity used.

The culverts are to be made of stone masonry laid in hydraulic cement, as are the aqueducts, except those where the streams are too near the bottom of the canal to allow space for turning the requisite arches. In such cases, wooden trunks are substituted on stone abutments and piers.

The abutments for bridges are of stone masonry, laid in quicklime mortar, at 50 feet apart, upon which a strong truss bridge is laid.

The bridges for roads much travelled are 22 feet wide. Other road and farm bridges 12 feet wide.

The plan for waste-weirs generally has stone abutments to sustain the earth, with a timber gate-way, and where the ground is favorably situated, it is designed to add long wastes, by a depression in the top of the bank paved with stone.

I have made estimates, in detail, for two plans of locks, with chambers 90 feet long and 15 feet wide: one for a composite lock

of wood and stone, like those now being built on the Chenango canal-and the other, for a lock of stone masonry, wholly laid in hydraulic cement. The quoin and coping to be cut stone; all the other masonry to be substantial and smoothly hammered work.

The composite lock from head to 8 feet below upper gates to be of stone masonry laid in cement-the remainder, a dry wall with a wooden chamber made within, bolted to and supported by the masonry. The plans for the foundation and gates are the same in both locks. For a detailed

Drawings of each plan of locks and of the aqueducts, bridges, culverts, &c., will be found at folios 32, 33, 34 and 35 of the Atlas. There is a difference in the expense of the same kind of structure, occasioned by the location and by the greater facility of obtaining materials in some cases than in others. The statements here referred to, however, are calculated from medium cases, and are designed to shew the average cost.

For the purpose of ascertaining the quality of stone, and the facility of procuring them, I examined the quarries within a reasonable distance of the canal. For the line from Genesee to Rochester, limestone will be used, and is found of a very superior quality near Avon Springs and at Caledonia. There is stone at Rochester which may be used in some parts of the work adjacent; but of an inferior character.

All the quarries from Mount Morris south, either on the Dansville side cut or in the vicinity of the main canal, are sand or grit stone, varying in hue from pale blue to ash color. They are fine building stone, (of the kind) quarry casily and in good shape, for the construction of the mechanical work; they are abundant and generally convenient to the line, especially on that part of the route from Portageville to Olcan.

It will be seen by referring to the detailed estimate of the different proposed kinds of locks, that the average cost of a composite lock of 10 feet lift will be $3,700.79, and of a hammered stone lock $5,193.98, making a difference in favor of the first description of lock of $1,493.19.

The difference between these two modes of construction is small when it is borne in mind, that the former will require a thorough repair of all the wood work connected with the chamber, once in 8 or 9 years, which is rendered still more expensive on account of the season of the year at which it would be necessary to perform the work, in order to obstruct the navigation as little as possible. The estimates throughout the line are made upon the latter mode of construction, which I think it would be well to adopt, at least between Mount Morris and Rochester, where the stone is particularly suitable, and where the suspension of the navigation would

be sensibly felt. On the other parts of the line it may be found advantageous to substitute the composite lock, as with the excep tion of a few quarries on Caneadea creek and in that vicinity, the stone is not of so good a quality as that below for making perma nent hammered work, coping or quoins.

FEEDERS.

The question of supplying the summit of this canal, and the lower levels dependent upon it with water, is one of the most interesting and important, which now presents itself for considera

tion.

On referring to gauges taken by Judge Geddes in 1825, on the same summit, I was led to doubt, whether a quantity adequate to the demand could be obtained, and consequently devoted much time to a careful investigation of the subject.

From the best information I could obtain, a more favorable op portunity has not occurred for several years, of ascertaining the quantity of water which may at all times be relied upon, for the supply of the summit level, on account of the very low state of the water in all the streams at the time of gauging. For this purpose dams were thrown across the several streams, with apertures and sluices, in a form best calculated to concentrate and shape the volume of water for measurement.

The following statement shows the minimum flow of running water, that may be commanded for the summit, viz:

To conduct this water to the summit 9 miles of artificial feeders will be required, viz:

Which, together with the requisite dams and bulkheads, are estimated to cost $29,069 33.

The water from these lakes must be turned into Ischua creek, and occupy its channel to Ischua feeder, a distance of about 14 miles, through which it will be received into the summit level.

Lime, Beaver, Mud and Fish lakes, lie in the towns of Machias, Freedom and Farmersville, all within the compass of 7 miles.

Lime lake is 18 miles from the canal, and 156 feet above the summit level. Beaver lake is 209, Mud lake 218, and Fish lake 2671 feet above the same point. The south end of Lime lake is separated from the Ischua, by a Tamerack swamp 11 miles in length, and rises on an average, 21 feet above the surface of the water in said lake. At the time of making the survey, I was informed that the water in this lake had been raised from 2 to 3 feet above its ordinary level, by a dam at the outlet, and that it was the intention of the proprietor, to raise it 2 feet more. This may be done at small expense, and by excavating a channel to the Ischua, not exceeding 8 feet in depth, the lake may be drawn down 7 feet.

On the outlet, near the above mentioned dam, a grist-mill is being built, estimated to cost $5,000, and should the water of Lime lake be turned to the south, this mill will sustain an injury and be a subject of claim for damages. The reservoir would average an area of 135 acres by 7 feet deep.

Fish lake is divided into two parts of unequal size and of different elevations; the upper and largest division is 6 feet above the other, and in the dry season has no discharge into the lower division, unless it be through a subterraneous passage.

Willard Law has a saw-mill at the outlet of the lower pond, and is building a dam 180 feet long, by which he designs raising the water so as to flow back into the upper pool 21 feet.

In order to use this lake as a reservoir, it is proposed to raise Mr. Law's dam 7 feet higher, which will give an area of 120 acres, from which 10 feet in depth may be drawn.

The outlet flows into Clear creek, a branch of Cattaraugus creek. Between Fish lake and Cattaraugus creek, there are 3 grist-mills, 3 saw-mills, and 2 card and clothing-works, which in part receive their supply of water from this source, and should the water be used for a feeder or reservoir, claims for damages will

ensue.

Mud lake lies about three miles south west of Fish lake, and upon the plan proposed, the waters of the latter may be drained 10 feet deep below its contemplated surface into the former, by a channel not exceeding 9 feet in depth and 2 miles in length. On referring to the map, it will be seen, that the waters of these lakes will then pass on through a natural channel, and unite with those of Beaver lake near its outlet, whence they will be conducted into one of the branches of the Ischua, by an artificial feeder of 2 miles and 66 chains in length. Mud and Beaver lakes cover 50 acres each, and may be conveniently raised 6 feet higher, which will add about 70 acres to their area=170 acres. On the outlets of these lakes, there are a few saw-mills and other works of snall note, which will sustain some injury The surface of country drained by them, is estimated at 20 square miles. Their principal and most permanent sources of supply, especially in the dry seasons, is from springs.

It is difficult to determine with much precision, the extent of damage that may be sustained by diverting the waters of these lakes from their natural channels. They flow north into Cattaraugus [Assem. No. 42.]

4

creek, and should there be any want of water at the mills on that stream, it may not be improper to suppose, that claims for damages will be presented; the most depressed of these lakes, is about 1,072 feet above Lake Erie.

The following tabular statement presents at one view, the capacity of the lakes, and other streams available on the summit level; the quantity received from the natural discharge of each stream, after deducting the leakage, filtration and evaporation, occasioned by conducting their waters through the several artificial channels to the summit, the quantity that may be held in reserve by damming the lakes to be used during the dry season, which from my investigations on that subject, I have reason to believe, rarely if ever exceeds 140 days, showing also how much may be drawn per minute for the above time, after making allowance for the filtration and evaporation incidental to the increased surface of the lakes, together with the cost of construction, which will perhaps, be as well calculated to aid you in determining the relative advantages and feasibility of this plan of supplying the canal thing I may say further on the subject.

with water, as any