Connection at Savanna was made by running in duplicate from old United States permanent bench marks 62, 63 and 60 to our T. B. M. 353. Thus three independent elevations of T. B. M. 353 were found after applying all corrections. They were quite concordant.

The mean of these three different elevations of T. B. M. 353 was taken for its true elevation=607.450 feet above Cairo datum plane. The elevation of this T. B. M., as carried down from Savannah by precise levels after applying rod correction 176.88604-580.340 feet. Constant to be added after all corrections are made = 8.26314m 27.11 feet.

The number of points, of which the elevation is computed is 696. The number of points described in the list of bench marks is 553.

All elevations are first determined in meters and then converted into feet.

No large error was found in the field reduction. One error of 10mm, one of 6mm, and one of 3mm, several of over 1mm, and quite a number with magnitude less than 1mm. The probable error of the final result is computed for each bench by obtaining the square root of the sum of the squares of the probable errors of each stretch intervening between it and the initial point. This, for the last bench in the line, gives the probable error to be 13.77mm

To find the probable error per kilometer, we assume the whole line of 484.68 kilometers to be divided into stretches of 1 kilometer in length. Then the probable error per kilometer will be such that 484.68 times its square will give the square of 13.77,.... 484.68x2=(13.77)2 .......... 22.015x=13.77. Probable error per kilometer =x=0.625mm

The value found for x for the first 50 kilometers was
The value found for x for the first 100 kilometers was
The value found for x for the second 100 kilometers was.
The value found for x for the third 100 kilometers was
The value found for x for the fourth 100 kilometers was.
The value found for x for the last 85 kilometers was.

+0.66

0.65

+0.66

0.61

0.58

0.60

Running very regularly and showing that all the work is of about the same grade of precision.

Recorders F. B. Williams and John P. Baker have assisted me in this final computation, and are to be commended for their rapidity, accuracy, and comprehension of the work.

The work of checking the conversion of meters to feet on Mr. Paige's tabulation from Duluth to St. Paul, and appending the elevations in meters and feet to his description, has also been done."

Respectfully submitted.

Capt. CARL F. PALFREY,

Secretary Mississippi River Commission.

O. W. FERGUSON,
U. S. Assistant Engineer.