GLASGOW.

This was found to be a counterbalanced iron-wire gauge on the north side of the river nd on the first deck-span approached from the south. The gauge wire is No. 10, B. V. G., and a weight of thirty pounds was attached to it. It was found reasonably orrect, but its location is such that during the lower stages of water the bar becomes ntirely dry underneath it.

The gauge is far from the main channel, and the only reason for placing it there ould have been to keep the counterpoise or gauge-weight out of the way of passing teamboats.

In low water the observer adopted the expedient of lowering a weight from a sheave n the top chord of the bridge.

A standard wire-cable gauge was placed over the main channel, and showed a differnce of .033 in height of water surface from that given by the old gauge. Observer Aris B. Sherwood, bridge attendant, is entirely satisfactory.

BOONVILLE.

This gauge, although found to be in very good order, and nearly correct as regards is zero, is open to the same objections which obtains to all gauges which are attached o draw-rests or any other obstruction to the free passage of water.

A standard cable gauge was established over the main channel on the draw-span f the bridge, and readings on both gauges ordered to be continued, in order to note he probable error of the vertical gauge.

Observer C. Randecker, watchman on the bridge, is fairly competent.

JEFFERSON CITY.

The gauges here are not very satisfactory. The high-water gauge is on the face of a ailroad culvert, spanning a good-sized creek, and its readings, when the creek is high, annot give the true height of water in the river. The low-water gauge is a temporary loping one, located about 1,800 feet above. It has served its purpose very well and robably will remain until next winter, but no absolute connection between the two auges can be made at different stages of water.

It is recommended that a new inclined gauge be placed in the locality of the lowvater one at an early day.

Observer D. W. Rogers, collector on ferry-boat, appears to be a very conscientious

aan.

HERMANN.

This is a Signal Service gauge, receiving no attention except that given it by its oberver, Mr. Maushund. It is made of 2-inch plank on edge, and nailed to small posts. t is quite well protected by large rocks, or it would have been swept out by ice or Irift. The lower end has been repeatedly broken off and been replaced by the observer. As this gauge has proven of benefit in investigation of physical data, it is recommended that a permanent inclined gauge be established here.

SAINT CHARLES.

This was the first standard wire-cable gauge established by your order on the Missouri River bridges. It was not inspected, but the observer, J. S. Hill, bridge watchman, a very careful and intelligent man, reports that he has kept a mark on the guardrail of the bridge which would show any variation in the length of the cable, and that no appreciable difference can be observed since the gauge was put up in March.

DESCRIPTION OF STANDARD WIRE-CABLE-BRIDGE GAUGE.

This is a very simple gauge, as the sketch herewith shows, and probably the most Teliable of all the gauges on the river.

The component parts are a deeply grooved iron sheave with a lag screw and washers to fasten it to a stringer, guard-rail end of tie, or any convenient timber; a sufficient length of best Swedish iron wire sash cable of inch diameter, running over the sheave, to one end of which is fastened a weight varying with the length of the cord, generally a window-sash weight of 14 to 16 pounds, and to the other end is attached an iron handle with an index cross-piece and a turnbuckle and lock-nut for taking up any stretch that may ensue from long use of the gauge. Next is a properly graduated gauge-rod fastened horizontally to the same stringer or guard-rail that the sheave is attached to, with its zero properly placed. In making an observation the observer simply lowers the weight until it touches the water and takes his reading. The contact with the water surface can be made quite accurately, especially if the sun casts a shadow from the weight.

8872 ENG 87——187

After the reading is taken the observer walks back with the entire line and weight until the latter is sufficiently high to be out of the way of passing steamboats or until it is up to the bottom chord of the bridge and secures the handle to a spike or stape and locks it thereto, if he deems it necessary.

It was at first thought that the size of the cable (†-inch) would present too great a wind surface, but it has been found that, even iù quite heavy winds, it does not become enough curved to vitiate an observation.

It is also found that the cable, while it is much stronger and more durable than iron or other wire, does not stretch, except to a very small extent, under the continue strain of a heavy weight.

Very respectfully, your obedient servant,

Lieut. T. A. BINGHAM,

Corps of Engineers,

Secretary Missouri River Commission.

A. H. BLAISDELL,

Assistant Engineer.

Appendix A6, showing stages of the Missouri River, from Saint Charles, Mo., to Fort Pierre, Dak., compiled from miscellaneous and regular gauge records, between 1872 and 1885, was printed in House Ex. Doc. No. 28, Forty-ninth Congress, second session, and is here omitted.

2979

APPENDIX A 7.

REPORT OF FIRST LIEUTENANT T. A. BINGHAM, CORPS OF ENGINEERS.

MISSOURI RIVER COMMISSION,

Saint Louis, Mo., April 1, 1886. SIR: In connection with the survey of the Missouri River, now under way, by direction of the Commission, question arose as to the reliability of the elevations determined in past years; because the instruments used had been wye levels with New York rods, whereas the so-called "precise" leveling of the Mississippi River Commission had been done with Kern instruments and self-reading rods. The Coast and Geodetic Survey people, who are engaged upon a transcontinental line of levels, have advanced as far west of Saint Louis as Etlah, Mo., a station on the Missouri Pacific Railroad. For nearly 20 miles (from Sonth Point to Etlah) their line lies along the right bank of the Missouri River where there are many of the U. S. E. bench-marks. It was deemed best to send a party into the field to level over this 20 miles for two objects, viz:

(1) By connecting the U. S. E. and C. S. bench-marks to get a check on our own elevations.

(2) To determine how closely to a line of "precise" levels a line can be run with a wye level, the Coast Survey method and that of the M. R. C. being substantially the same.

The party started February 8, and returned March 6. The work cost $365, not including salaries of assistant engineer and rodman, who were taken from the office force. Mr. Warren's report is appended.

While the party was absent in the field it was discovered that P. B. M. 12, of the Mississippi River Commission, at the mouth of the Missouri River, on which all the U. S. E. elevations depend, had not been finally determined when elevations were printed. Taking the figures given in Mississippi River Commission Annual Report for 1883, the correction to be applied to all our elevations is + 0'.341.

The accompanying diagrams were prepared by my direction to illustrate different methods of comparison.

Referring to sheet 1 and regarding the connections between the U. S. E. and C. s. benches, as accomplished without error, column 21 gives the differences between our corrected elevations and the Coast Survey elevations for the same bench-marks.

10mm dist. in kilometers (or 0.041621 √ dist. in miles)

is the limit of discrepancy allowed in the Lake Survey (see Prof. Pap. Corps of Engineers, No. 24, p. 598).

In the instructions for precise leveling, prepared by Mr. L. L. Wheeler, under the direction of General Comstock (Report Chief of Engineers for 1880, p. 2433), the limit of allowable discrepancy was given as

5mm dist. in kilometers (or 0.02081 √ dist. in miles.)

In the Annual Report Mississippi River Commission for 1883, p. 143, will be found the deduction, by Mr. L. L. Wheeler, that cumulative errors are "nearly if not quite proportional to the distance leveled, instead of the dist."

In Appendix 11, Coast Survey Report for 1882, p. 522, is found the following note: "With respect to leveling of precision executed of late years in Europe, in the opinion of the Geodetic Association a probable error in the resulting difference of height as large as 3mm per kilometer (0.01584 per mile) may still be tolerated, but one of 5 mm (0.0264 per mile) would be considered as surpassing an allowable limit. A value of 2 mm (0.01056 per mile) may be considered to represent a fair measure, and one of 1 mm (0.00528 per mile) a measure of high precision."

The severest of these limits is the second given, which is the one adopted by the Mississippi River Commission.

Applying these tests to column 21, we find that all but one of the differences are within the severest limit.

Since the U. S. E. levels were carried down the river from Sioux City, differences shown in column 21 are quite satisfactory.

There is one difference that does not fall in line, but uncertainty exists as to the exact bench-mark in this case.

It seems as if a constant error appeared in these differences as they stand. This constant is not yet satisfactorily explained. It may never be, but it is proposed to