Lake Erie into the Blackwell Canal, which attempt proved a disastrous failure, and, as I think, simply because it was not made in the right manner and with the proper precautions. The site is indicated on the appended sketch, and all that remains to show it is the double row of piles there indicated. It appears that an excavation was actually made 100 feet wide, and to a depth of nearly 14 feet below water level, from Lake Erie to the Blackwell Canal, the lake being separated from the canal by a temporary dam. A gale arose which breached the dam and filled the cut with sand up to the general level of the beach, whereupon the enterprise was abandoned, and which disaster, I presume, contributes to a considerable extent to the general distrust of further projects of like nature at that point. From this disaster we learn, as a little forethought might have previously shown to the projectors of the work above referred to, that it is useless to attempt to make a channel from Lake Erie into Buffalo Creek without previously protecting effectually the Lake Erie entrance from the effects of gales, and this can only be done by first building the protecting piers which must at any rate eventually be constructed, being a very essential feature of the entire project.

It is, perhaps, as well that this project did fail, inasmuch as the cut was to have been made so nearly at right angles to the axis of the Blackwell Canal at the point of entrance, that it would have been almost impossible for vessels of such lengths as we have at present to have passed in or out.

The next project I find, which still retains this faulty position of the cut, proposes to defend the Lake Erie entrance by a pier running the same course as the south United States pier at the mouth of the creek, and extending into the lake to 20 feet soundings; also to mark the channel limit on the north side by a short pier of less width. This device, while effectual in protecting the Lake Erie entrance, at the same time, by reason of the abrupt turn it necessitates at that point, makes the entrance quite as impracticable as the exit at the other end due to the erroneous course of the channel which it still retained.

The direction of this pier is indicated on the sketch in dotted blue, although not its full length. I have made use of the idea to suggest a plan which might be used in case a great saving in the items of piering might be desirable, even at the expense of convenience of ingress and egress. A single pier, as indicated on the sketch, with the north limit of the channel defined and preserved by a close row of piles, would be the least expensive and yet effectual protection to the Lake Erie entrance. But, at the same time, the channel would be inconvenient of access to all vessels, and almost impracticable to those of great length. I therefore reject this plan, excepting as a last expedient, for the reason that I consider it important that a work of this nature should be so constructed as to meet the wants of commerce fully for all time, so far as we can judge of what those wants will be in the future.

I now pass to my own experience and researches as to the best direction and construction of this channel. Upon his assuming charge of the works of improvement of Buffalo Harbor, Brevet Major General T. J. Cram, colonel of engineers, directed me, among other duties assigned to me as his assistant at Buffalo, to make a careful examination and survey of the ground in the vicinity of the site of the old project for the south channel, and to select and submit to him a site which should fulfill the conditions of necessitating the minimum amount of excavation and yet have the axis of the proposed channel intersect that of Buffalo Creek at

an angle convenient for the passage to and fro of vessels of the greatest length now on the lakes.

This task was fulfilled in the fall of 1867, and the result submitted with a survey, the essential part of which is reproduced on the inclosed sketch. The ground within the blue lines was reported as fulfilling the prescribed conditions to the best advantage. Upon relieving General Cram in charge of the works at Buffalo in the spring of the present year, I received from him plans, specifications, and bills of materials for a proposed south channel based on my survey of last year.

The project indicated by his plans and specifications is shown in full blue lines, and to this project, after full consideration, I dissent, for reasons which I will proceed to set forth:

1st. As to the direction of the piers. This is not laid down on General Cram's drawing, but it is inferred from their lengths that they were intended to be placed as indicated in the appended sketch. Any other position not open to the objection I am about to make would be open to the one already made, of making the turn at the Lake Erie entrance too acute. Discussing, therefore, the direction of the piers as indicated in full blue, there is first to be borne in mind the object of piers at the Lake Erie entrance, viz: to prevent the choking up of that entrance by sand brought along by the littoral current and thrown shoreward by the seas resulting from westerly gales.

Will the direction of the piers S' and N' prevent this from taking place? I say no. At the first start there is no cause for apprehension, because the littoral current bringing down the sand in suspension will deposit it in the angle A. But when that angle is filled to the level of the beach, which will not be very long, (judging from what happened to the channel when it was destroyed,) and when the shore makes, as it will eventually, in a curve out to the very end of the south pier S', how can the 14 feet of water be preserved at that point, and what is to prevent the channel from choking up, as in the case of the old one, on the first heavy gale after the beach reaches the pier-head? The direction of the heaviest seas, determined by personal observation on my part, is laid down on the sketch. A comparison of this direction with the course of the piers will illustrate my course of reasoning.

This course of reasoning led me to the conclusion that no system of parallel piers would answer in this locality, unless inclined to the northward of the general direction of heavy seas, which would make the objectionable abrupt turn at the lake entrance of the cut, and even then the south pier must be made much longer than the north one in order to force the bar, which must inevitably form at some time or other, to make at a sufficient distance from the north pier-head to admit of a channel remaining between that pier-head and the probable bar sufficient for vessels to pass until the bar could be dredged away. But this gives another sharp turn at the north pier-head, and taken altogether, a very inconvenient, indeed almost inaccessible channel.

I advance, therefore, the following propositions:

1st. That the littoral current holding the sand in suspension, would form either a beach or a shoal against the south pier, no matter what its course. 2d. That the beach or shoal when made would constitute a magazine, from which in gales of wind sufficient sand would be drawn in one gale to form a serious bar at the head of that pier. 3d. That in the case of the two parallel piers S' and N', this bar will almost, if not quite, fill the channel and render it inaccessible for a time until it shall have been removed, at a great expense sufficient, perhaps, to counterbalance the advantage of having a channel at this point, in which case

that channel would meet the fate of the old attempt before referred to. That disastrous project might, perhaps, be adduced as sufficient argument to condemn the project of two parallel piers only, as a defense to the Lake Erie entrance; but in order to elucidate the subject fully, I have to request attention to the appended sketch, on which I have indicated the probable bars, marked B.

Now I argue that these deposits will be comparatively trivial in all cases but one, and that the plan indicated in blue and its similar cases. In that case the waves having thrown the sand in suspension it will be borne along by the littoral current, and finding the cut as indicated in blue, with open arms ready to receive it, will be forced into that cut, and, I venture to say, but little sand will pass the cut until the gale subsides. This of course will necessitate a formidable amount of dredging after each gale, and in my opinion more than could be profitably done. In fact I think that the cut under such circumstances would have to be abandoned.

Therefore, after due thought and mature deliberation, I have projected the plan indicated in red, which is respectfully submitted as best combining the three essentials, economy, accessibility, and durability.

1. Economy. On this score as well as that of accessibility I have selected the site of the old cut for the lake entrance. To carry it further to the southward would make the curve of entrance too abrupt, and bring us on no lower ground. To carry it further to the northward would bring us on higher ground and necessitate more excavation.

On the site of the old cut where I have located the entrance I gain an easy curve, and beside, the advantage of the old excavation, which is filled in with loose sand only, and lies almost entirely within my project, thus saving me just that much excavation in clay.

After leaving the Lake Erie entrance I take as nearly as possible the route of my survey of 1867, which follows the lowest ground. I incline slightly to the northward in order to avoid the limit of outcropping rock in Buffalo Creek, so as to get rid of the contingency of blasting.

For the rest on this point I have the honor to call attention to the comparative estimates herewith appended drawn between my plan and that indicated by General Cram's plan and specifications, by which it appears that my project is much the cheaper of the two.

2. Accessibility.-Beginning with the Lake Erie entrance I have to invite attention to sketch A, which shows the construction which I havė adopted.

To avoid the difficulties heretofore discussed, I run out my parallel piers at right angles to the beach, which position is the most natural, gives the least amount of dredging between the piers, and forms the most effectual barrier to the sand borne along by the littoral current. I then run off from the south pier, at an angle to the direction of heavy seas, a prolongation Ss, the end of which resting in about 20 feet water, shall be 500 feet from the north pier head and overlapped as to the course of heaviest seas nearly 200 feet. I thus give smooth water at the mouth of the cut, and allowing for the formation of the probable bar, will, according to my observations at the present mouth of the creek, even after the severest gale, have at least 350 feet channel between the bar and the north pier head. This bar will only form in the interval of time between the subsidence of the sea and the ultimate settling of the sand in suspension; for while the sea is up, striking against the pier prolongation Ss it is deviated in a course opposed to the direction of the littoral current, and will drive the sand in suspension back to the edge of the shoal or beach, as the case may be, at the salient Q. It is only when the

force of the incoming sea has ceased to act, that the littoral current will again resume its course. Thus the very force, which in the project I have already objected to would tend to fill the cut with sand, I make use of to diminish the extent of the inevitable bar. Two hundred feet is the minimum width of a straight reach of channel for the proper accommodation of modern lake shipping. Where curves occur this width should be increased more or less, according to the greater or less degrees of curvature.

From the 350 feet wide entrance between the shoal and the north pier head the channel narrows down to 220 feet in the straight reach, the 20 feet over the 200 being necessary to allow 200 feet in the jaws of the channel at the junction with the beach. Thence I widen out toward Buffalo Creek in order to give 250 feet radius to the tangent curve at the next bend, which is about the minimum admissible for the accommodation of vessels of such lengths as we have at present on the lakes. From this curve I propose to narrow down again to 200 feet in width at the Buffalo Creek entrance. This course is respectfully submitted as the best as to accessibility that the circumstances of the case will allow, still having due regard to economy.

3. Durability.-As regards this, I place my principal dependence on the prolongations Ss, which will defend the entrance from the inroads of the sand, and keep the channel serviceable. I propose to construct the prolongation of substantial crib work, varying in width from 20 feet at Q to 30 feet at the pier head. For the remainder of the channel, I propose to adopt the construction indicated by General Cram in his plans and specifications, excepting the pier N, which being in the case of my project entirely protected by the pier S, Ss. I do not consider it necessary to give it a greater width than 12 feet, that of the canal diking. The mode of construction proposed by General Cram, while more economical than crib work, can be used to good advantage in the less exposed positions, being especially suited to the particular locality, the soil being, as before stated, a stiff clay overlaid by a thin stratum of sand not more than eight inches in thickness at any point. It has been urged in opposition to this channel that it could never be maintained even when constructed; that the dredging required to keep it open would be so great as to swallow up in its cost the advantages to be gained from having the channel. This might, indeed, be feared were the excavation all, or principally in sand or other loose soil. But, being principally in clay, we have nothing to fear excepting from the accumulation of the sand brought along by the littoral current, and this I claim to have fully provided for by interposing the pier prolongation Ss at the Lake Erie entrance, which I expect to prevent the ingress of any of the Lake Erie sand further than the bar B.

The objection which has also been made, that the proposed cut would damage the shipping interests of the lower interior harbor, may also be answered in a similar way.

In fair weather there is no appreciable current in Buffalo Creek at the point where it is proposed to cut the channel, the water in it and Lake Erie standing practically at the same level. I therefore do not anticipate any current one way or the other in the channel after it is constructed, excepting when the water of Lake Erie or Buffalo Creek is disturbed by some extraneous cause. The principal disturbing cause in Lake Erie will be a westerly gale; in Buffalo Creek a freshet from the interior. Let us examine the effects by referring to those which now occur under like circumstances. Westerly gales pile up the water of Lake Erie against the Buffalo shore, flood the low land in the vicinity of the proposed channel, when the water subsiding passes off, back into the Lake,

some as it came, but a large volume is discharged through the medium of the channel of Buffalo Creek. The action of this water passing off, is of course somewhat the same as that of a freshet from the interior, and the two may be discussed together. In each case, of course, there results a deposit from the flood, made at different points in the lower interior harbor. In the one case the sand swept over from the lake by the flood; in the other the detritus brought down from the upper creek. But will this deposit be increased by the presence of the cut? I think not. There will be no current through it, as the action of the westerly gale will be as it is now, to back up the waters of Buffalo Creek wherever the water in Lake Erie in its rise meets the waters of the creek. The water forced through the cut meeting then with the water forced up the creek from its mouth, will be effectually stayed and the deposit, if any, will be in the cut itself or up the creek, but to no greater extent than would have occurred had not the cut been constructed, inasmuch as the only increment to the deposit which might result from the existence of the cut would be the sand of Lake Erie borne along by the littoral current, and sucked in by the current in the cut, which current. I trust I have demonstrated will have no existence.

If this be admitted, all the objections as to damage of the lower harbor fall to the ground.

But on the other hand there is an effect on the cut itself of the downward current in Buffalo Creek, resulting from subsiding floods or freshets in the interior. The Buffalo Creek entrance of the cut being on the concave bank, the outsweeping current will probably strike full into it and deposit a formidable bar in the cut and Blackwell Canal. Should this be found to be the case, it will be necessary to construct a wing dam, D, to prevent it. Or a still better plan could be adopted, having in view the interests of the up-creek commerce, which will become of importance as soon as the channel is made. The cut as I have planned it, to accommodate the down-stream shipping, has its Buffalo Creek entrance of difficult access to vessels from up the creek, which are but few in number at present, principally barges from the iron works, &c., above. But should this shipping increase in number, as would probably be the case were the cut made, it might be accommodated by diverging the southern canal dike at C, and dispensing with the wing dam D, building an axial pier to the creek, starting from the Ohio Street Bridge and extending about 400 feet; leaving a channel of about 100 feet between it and the east bank, which is the one now used. This would render the cut accessible to up-stream shipping, and yet not render it liable to damage from the effect of freshets, for the pier P, would, I think, answer all the purposes of the wing dam D, and better too.

It will be impossible to prevent, entirely, the deposit of sediment in the cut during a freshet; but either the wing dam or pier would materially diminish the amount, by deflecting the current from the mouth of the cut, which current would carry the bulk of the sediment down the creek. There would still exist, of course, a current from creek to lake through the cut, due to the difference of level of the water in the creek and lake, but this, from the uniformly varied nature of the force, would have a uniformly varied velocity, and consequently deposit the sediment uniformly throughout the cut, excepting, perhaps, at the points b+b', where small bars might be formed.

But I do not anticipate a sufficient deposit, even including the small bars, to necessitate more dredging than is required now, annually, at the mouth of the creek. If it should be objected that the position of the pier would narrow the channel too much, the answer is that it would simply define the channel as it now exists, the bottom of the creek being