unusal course also produced the erratic phenomenon of a “backing" wind holding steadily from one direction (sco Toledo record) for forty hours. Many disastrous wrecks occurred, and it is worthy of note that several of the worst were in the vicinity of the "Narrows," before mentioned, between Long Point and the American shore near Erie and Dunkirk. The steamers Dean Richmond and Wocokea and the schooners C. B. Benson Riverside were all lost during this storm, and all in the same locality-at these "Narrows." Seven lives were lost with each of the schooners, which were considered to be seaworthy boats, as both were in the grain trade. Not a soul was saved from the Richmond and but three from the Wocokea. CONCLUSIONS. It will be observed that while few of the heights are accurate, they carry sufficient reliability to warrant a general discussion of the matter in the light of this crude data, and the hope that it will lead to the obtaining of more definite knowledge and perhaps more sound conclusions. A discussion of other points of interest than those touched upon in this report, though tempting, is, I feel, hardly justified by the data now at hand. It must also be noted that all data we have was recorded along the south shore of the lake and that the times of record are not coincident, though generally nearly so. Following now the profile of water and surface in connection with the above data, we find that in the West Basin the fall in the funnel-shaped end, containing Monroe and Toledo, was 6.8 feet; in the open it was 5.3 feet. In the main basin, immediately we pass the Island barrier, the fall was but 2.6 feet for all points until Cleveland is reached at the widest part of the lake. Here we find a fall of but 1.2 feet, and practically the same at Ashtabula. Between Ashtabula and Conneaut, a distance of 13 miles, we meet a solid wall of water 4.7 feet high, there having been a rise of 3.4 feet at Conneaut. At Eric the rise was 0.8 foot less than at Conneaut, and at Buffalo the highest point reached was 5.3 feet. The question now at once arises, do these surface heights along the south shore correctly represent the heights of water in the lake? Assuming it to be so, that is, that the surface is level on each line normal to the shore, then the surplus water in the eastern end of the lake should be about equal to the deficiency in the west end. We find the line of no variation from the normal stage before and after the storm to have heen between Ashtabula and Conneaut. The area west of this line is approximately 7,000 square miles, that cast of it, 3,000 square miles. In order to make the two quantities equal, the proportion of fall to rise should be as 3 to 7. The record of fall being more numerous, covering larger territory and to a fair extent agreeing among themselves, we may assume a fall of 5.3 feet over 1,200 square miles, 2.6 feet over 1,800 square miles, 1.2 feet over 4,000 square miles, to equal à fall of 2.3 feet over 7,000 square miles; which would give, if our assumption is correct, an average rise of of 2.3 feet 5.4 over 3,000 square miles. This is not borne out by the data, as it is as large as the maximum height at Buffalo, and twice as large as that at Erie. To show its absurdity, we have really an average rise of perhaps 3.7 feet over 3,000 squaro miles, which leaves unaccounted for a body of water amounting to 16,000,000,000 of cubic feet, enough to supply the ordinary outflow of Niagara for 20 hours. This amount is beyond that already accounted for by the recorded rise at Buffalo, which by itself would scarcely double the outflow even while it lasted. When we consider the pressure which must have existed in connection with the change of elevation of 4.7 feet in 13 miles between Ashtabula and Conneaut, and consider also the immense volume of water displaced west of them and not found to the eastward, it suggests the idea of an enormous eddy or swirl, more or less forcible, in that portion of the lake, the current setting down along the south shore and up along the Canadian side. The observations, though crude, seem to show conclusively that the surface of the lake is not level on the normal lines, but is much higher on the north shore. This is borne out by the fact before mentioned that during the height of the gale there is invariably a reflex current into the west point of the lake, too strong to allow of the belief that it is caused by the gravity of the "piled up" water overcoming the force of the wind. It is a fact that most of the Lake Erie wrecks during a westerly gale are in the vicinity of the "Narrows." In an easterly gale, when these peculiar conditions do not exist, there is seldom a loss in that region. Is it not possible that in these serious storms there are forces at work for destruction with which we are not familiar, and that a proper study would give the means for combatting them successfully? I beg leave to suggest that the line of thought here touched upon is important enough to warrant further and more accurate study on the basis of definite data. If the conditions here suggested do actually exist, it will be of vast importance to the navigation interests to know of them; and in order to gain the information for a more thorough and accurate discussion, I respectfully recommend that steps be taken to have all light-keepers on Lake Erie record the height of water three times daily, together with the wind directions and velocities, and to have special instructions issued that hey shall note any unusual conditions of weather or water. In any prolonged and severe storm, particularly those of April and October, they shall carefully note the extremes of water level and wind. To be of any service in a further consideration of this subject, the Canadian authorities should be requested, through the proper channels, to cooperate fully. I am thoroughly impressed with the idea that such observations will bear fruit of some importance at very slight expense. Very respectfully Lient. Col. JARED A. SMITH, WM. T. BLUNT, Corps of Engineers, U. S. A. WATER LEVEL OF LAKE MICHICAN. REPORT OF MAJ. JAMES F. GREGORY, CORPS OF ENGINEERS, FOR THE FISCAL YEAR ENDING JUNE 30, 1894. UNITED STATES ENGINEER OFFICE, * GENERAL: I have the honor to forward the accompanying plate on which is continued the water level curve on Lake Michigan for the fiscal year ending June 30, 1891, and to inclose a letter from Lieut. C. H. McKinstry, Corps of Engineers, giving the monthly mean water levels during the year. Very respectfully, your obedient servant, Brig. Gen. THOMAS L. CASEY, Chief of Engineers, U. S. A. JAMES F. GREGORY, Major of Engineers. REPORT OF LIEUT. C. H. M'KINSTRY, CORPS OF ENGINEERS, UNITED STATES ENGINEER OFFICE, SIR: I have the honor to transmit herewith the water-level curves* for Lake Michigan for the year 1893-'94, from tridaily observations taken at Milwaukee, Wis., and Escanaba, Mich, with monthly reports of observers.* Observations at Milwaukee were taken continuously throughout the year; at Escanaba they were discontinued from December 17, 1893, to March 18, 1894. Following are the monthly means (feet and decimals below plane of reference) from which the curves were plotted, the plane of reference being "high water of 1838." Stations. July. Aug. Sept. Oct. Nov. Dec. Jan. Feb. Mar. Apr. May, June. The "reduction to the plane of reference" at Milwaukee is 0.61 foot; that is, the zero of the gauge is 0.61 foot above the plane of reference. "reduction to the plane of reference" was determined in 1877 * Omitted. At Escanaba the to be -0.76 foot (B port of 1876, Vol. I, p. 84; 1877, Vol. II, p. 1194), and the monthly means from that time until June, 1882, were corrected (reduced) by that amount. The observations from July, 1882, to June, 1887, were further reduced by 0.187 foot. (Report of 1887, p. 2417.) The observations from July, 1887, to June, 1892 (published in Report of 1892, p. 3430), and the observations for 1892-93 were reduced by 0.76 only. This was plainly an oversight, and I would respectfully suggest the desirability of making a further correction of — 0.187. In July, 1893, the zero at Escanaba was tested by leveling from bench marks in the vicinity and was found to be 0.902 foot above the plane of reference. This correction (0.902 foot) was used in reducing the observations of 1893-'94. Very respectfully, your obedient servant, Maj. JAMES F. GREGORY, Corps of Engineers, U. S. A. C. H. MCKINSTRY, First Lieutenant of Engineers. WATER LEVEL OF LAKE ONTARIO. REPORT OF CAPT. DAN C. KINGMAN, CORPS OF ENGINEERS, FOR THE FISCAL YEAR ENDING JUNE 30, 1894. (For letter of transmittal see Appendix PP.) Permanent gauges are established at Oswego Harbor and at Charlotte Harbor (at the mouth of the Genesee River), and each has been read three times per day during the year. They show the lake level to have been lower than usual at similar dates throughout the year. OSWEGO GAUGE. This gauge was established in 1837 by the United States Engineer at Oswego at plane of extreme low water. The lake level has several times since been at this plane, but never below it. The gauge is cut on the harbor face of the stone pier at the foot of West Third street, and is indıcated by an iron plate cut to feet and tenths and bolted beside it. The zero of the gauge is referred to the top of an iron bolt in top of masonry of old Government stone pier 0.5 foot from east face of pier, 33 feet north of its intersection by the crib work wharf, foot of the United States reservation at the foot of West Third street, Oswego, marked U. S. B. M. This bench mark is 7.75 feet above zero of gauge. The zero of gauge, on plane of extreme low water, is 244.21 feet above mean tide at New York. (See p. 609, Prof. Papers 24.) Readings were taken daily at 7 a. m., 1 p. m., and 6 p. m. with observations of direction and force of wind. The daily means were taken, and a mean of three taken as a monthly mean. |