The following text may have been generated by Optical Character Recognition, with varying degrees of accuracy. Reader beware!

a MARINE REVIEW. oy feet, and for mean low water 8.6 feet, making the mean low water slope greater than that at high water. In extreme cases, the slope at the low water stage exceeds that at high water by nearly one foot, and since the discharge of any river is a function of its cross-sectlon and slope, we have here the anomalous case of a river sometimes having its maximum dis- charge at its lowest stage, This is an important condition, which has not been given consideration in previous discussions of this problem, and which is really the fundamental principle which should be observed in making investigations to determine the river discharge, and in developing - plans for the improvement of the waterway. The level of Lake Erie, after reaching its maximum stage, commences to fall about one month earlier than Lake Huron and about three months before Lake Superior, showing that these lakes, when acting in their max- imum natural capacity as reservoirs, are inadequate to maintain the level of the lake into which they discharge. Lake Michigan commences to fall at the time the maximum discharge is taking place in the St. Mary's River, and sometimes falls as much as eight inches per month, which would correspond to an actual discharge ot 320,000 cubic feet per second in excess of the inflow from the tributary drainage basins, making a total of over 500,000 cubic feet per second actually taken from the volume of water in Lake Michigan and Lake Huron. Evaporation from the lake surfaces during dry favorable seasons, is probably at times as much as six inches per month, and in connection with the rapidly increasing slope in the St. Clair River, constitutes the principal cause of the rapid fall of the water levels of Lake Michigan and Lake Huron. These sudden changes of level, corresponding to a rate of discharge of over 500,000 cubic feet per second, seem to indicate that any system of reservoirs which may be devised by damming Lake Superior, or otherwise, will be wholly inadequate to maintain the levels of the lower lakes ata stage very much higher than under existing conditions, and since the vari- ation in the slope of the St. Clair and Detroit Rivers seldom amounts to one foot, it would seem that the natural solution of the problem would consist in maintaining the surface of Lake Erie at as nearly a constant level as possible. An accurate knowledge of the outflow of the lakes is absolutely essen- tial in designing plans for regulating the lake levels. Unfortunately, such data do not exist at present, and from some inexplicable reason, the depart- ments having the authority to make such determinations, have been very adverse to making any observation to clear up the uncertainty. In 1841 Messrs. Z. Allen and E. R. Blackwell made a few observations near Black Rock with surface floats for the discharge of Niagara River. These observations, corrected from soundings made by the United States Lake Survey, give a discharge of about 250,000 cubic feet per second for mean level of Lake Erie. In 1867, 1868 and 1869, parties under the direction of Mr. D. F. Henry, Assistant United States Lake Survey, made three long series of observations with double floats and current meters to determine the discharge of the St. Mary's, St. Clair, Niagara and the St. Lawrence Rivers. The mean of all these observations gave for the St. Mary's River, 86,000; for the St. Clair, 225,000; for the Niagara, 265,000; and for the St. Lawrence, 300,000 cubic feet per second for a mean stage of water in the respective lakes. In 1891 and 1892 the United States Engineer Corps made two series of observations just below the international bridge in the Niagara River, and obtained for a mean stage of water a discharge of 230,000 cubic feet per second. In the report of this work the authors publish two formulas for com- puting the discharge, which give results so widely discrepant that no reli- ance can be placed on either. The difference of 35,000 cub.c feet per second between the results of these determinations make it quite evident that more reliable data must be had before the details of plans and esti- mate of cost for regulating the lake levels can be definitely stated. A comparison of the annual rainfall on the lake basins, with the respective outflows, shows that the mean discharge is a little less than one- nalf the mean annual rainfall on the tributary basin, and about one cubic foot per second for each square mile of the basin. In the observations made to determine the discharge through the St. Clair river, the results were reduced for that of mean stage of Lake Huron, without regard to the existing slope of the river at the time the observa- tions were made. Since, as has been shown, the outflow of Lake Huron de- pends on the lake level, and also on the difference in level between Lake Huron and Lake Erie, it is somewhat questionable whether the corrections for reducing the result to that of mean lake level were correctly applied. The action of strong winds on the lake surface often changes the levels from two to three feet; and unless these changes are kuown, and the proper correction made for the effect on the river discharge, large errors and discrepancies will, very likely, be foutid in the results. That is, while the observations may be correct for the day on which made, the result for the same stages of water obtained on some other day may be widely different, The discharge of the St. Mary's, the Niagara and the St. Lawrence rivers depends entirely on the stage of water in the lakes from which it flows, the determination of which is, therefore, a much more simple problem than for the St. Clair, The plans and methods for improving lake harbors, inaugurated up- wards of seventy years ago, are still in vogue; but, judging from the annual complaints in regard to the deterioration of channels at harbor entrances, it is evident that the limiting depths have been reached where such methods can produce permanent results. The general plan has been to deepen the channels of connecting water- ways by dredging, and at the entrance of harbors to construct parallel piers 150 to 200 feet apart, and dredge between. This system for depths of less than twelve feet was very effective, but far greater depths the annual silt- ing up of channels makes the cost of maintenance very great. 'The width between piers is also much too small for the depth of twenty feet or more, which must soon be obtained for all of the principal lake ports. The goy- ernment has endeavored to maintain a depth of seventeen feet at these ports for several years; yet, at the opening of navigation for 1895, owing to the low stage of water in the lakes, and to the silting up of channels, but few of the harbors (except on Lake Superior) had navigable channels of over twelve feet. The movement of sand parallel with shores during severe storms is enormous, and wherever these harbor channels have been excavated much deeper than the adjacent bottom of lake, silting is almost certain to occur. _ In the new channel, recently cut at the entrance of Toledo harbor, the government engineer in charge reports an annual filling up of 0.5 feet; or assuming the average-depth of the cut to be eight feet, the work of main- tenan e amounts to the same as for cutting a new channel every sixteen years. When we take into consideration the fluctuations of the lake levels, it is evident that a channel 21 feet deep at mean lake level will not generally insure safe navigation for vessels of 20 feet draft, except for a period of about three months each year. At the opening and closing of navigation, low water usually prevails, and vessels constructed for a full load on a oC foot draft will be able to carry only a part of what would be an economical load in a waterway regulated to fixed level. : The total fluctuations of the lake surfaces being over four feet, a well devised system of regulation, which will reduce the range of levels to one foot or less, will increase the low water depth of all the channels three feet, - and fix the draft for which vessels may be constructed so as to carry full loads throughout the entire season of navigation. Two different methods have been advocated for accomplishing this re- sult: First: To make a regulating reservoir of Lake Superior and its tributaries, and turn their surplus waters into ake Huron at the proper season, to maintain the low water stage at a higher level than under the present regimen ; and, second, to regulate the level of Lake Erie with a submerged dam near the outlet of the lake, such as to increase the low- water stage about three feet, reduce the fluctuation to less than one foot, and indirectly through the change of slope in the St, Clair River, reduce the fluctuations of the water surface of Lakes Huron and Michigan to at least one-half that at present. With the data now at our disposal, it is impossible 5 state definitely what may be accomplished by either of these plans, or what the exact cost would be to complete them; but the physical facts are sufficiently well known to be able to point out in a general way the results which may be reasonably expected from the execution of either one or both of the plans, and to conclude definitely whether such an undertaking is worthy of an official investigation. The extreme fluctuation of the water surface of Lake Superior is about 3.5 feet, with a mean fluctuation of only 1.2 feet. With proper regulating works near the foot of the lake, the low water stage might be maintained at least one foot above the present mean level, without damage to any ex- isting structures. The extreme high stage of the lake is approximately two feet above mean lake level, and to avoid all litigation for damage from overflow, the regulation should be such as to prevent this limit from being exceeded. The present outflow for mean level of lake is about 86,000 cubic feet per second, and with regulating works completed, the outflow could probably be increased to 150,000 cubic feet per second for a period of three or four months each year, provided that such a discharge did not create currents in the St. Mary's River of such a nature as to make navigation difficult and dangerous. This amount could not safely be exceeded, and it is very likely the requirements of navigation would limit the maximum safe outflow at a much smaller amount. Lake Nepissing could also be © utilized as a regulating reservoir, but the supply from this source would be small compared with that from Lake Superior, and it would not, therefore, be safe to estimate on a total of over 75,000 cubic feet per second in excess of the present mean discharge of 86,000 cubic feet. Since the evaporation and outflow frcm Lakes Michigan and Huron sometimes amount to over. 300,000 cubic feet per second in excess of the influx from rivers and drain- age, and that too when the discharge through the St. Mary's River is a maximum, it is quite evident that the supply from the regulating dams alone would have but little effect in checking the fall of the water surfaces of the lower lakes. It has already been shown that the discharge of the St. Clair River varies rapidly with any change of slope, and since the mainte- nance of a high level in Lake Huron when that of Lake Erie is falling will increase this slope, the St. Clair discharge would be correspondingly in- [Continued on page 11.] shin a esd taae