evaluation can be placed at the present time, include the adjustment of the interterritorial freight rate situation and rate policies of the railroads with regard to water competition. In order to take into account all of the above factors, of which no estimate can be made, an arbitrary 25 percent of the total traffic was added to the tonnage developed in the survey. A unit saving equal to the average saving was applied to this arbitrary additional tonnage.

75. Amount of prospective commerce and savings.-The commerce considered immediately prospective for the proposed waterway and the savings which would accrue to the general public in reduced transportation charges, based on freight rate levels of July 1, 1938, are given in the following table, which also shows the commodities and the waterway ton-miles:

TABLE 19.-Prospective commerce and savings

76. Principal commodities.-Arranged in order of volume, the six leading commodities up-bound and down-bound are shown in the following table. These commodities comprise 84.8 percent of the up-bound, 91.1 percent of the down-bound, and 87.4 percent of the total commerce credited to the waterway.

TABLE 20.-Principal commodities in order of volume, Tombigbee route

77. Summary of commerce credited to waterway.-A summary of the tonnage, value, savings per ton, and average waterway haul of each class of commodities, is given in the following table:

TABLE 21.-Summary of prospective commerce, Tombigbee route

78. An analysis of the above tables reveals that the tonnage movements would not be balanced. The up-bound movement would be approximately 1.4 times the down-bound. There is also a great disparity between the up-bound and down-bound average saving per ton, the up-bound being 2.3 times the down-bound. The unbalanced tonnage movement indicated by the traffic survey is attributable to the large volume of petroleum products moving from the Gulf to inland points, principally in the Tennessee Valley. Petroleum products account for 42 percent of the up-bound traffic. The saving on this commodity of $2.14 per ton being relatively larger than on most of the other commodities results in an average saving per ton up-bound much higher than the average saving per ton down-bound. The downbound traffic contains no large movements on which relatively high savings are indicated. The largest single movement of logs shows an average saving of only 66 cents per ton, while the other large item, sand and gravel, shows a saving of only 22 cents per ton.

79. The average haul up-bound, shown in table 21, is longer than the average haul for down-bound traffic as a result of the longer movement of petroleum products. The large items of down-bound traffic move only a relatively short distance.

80. Analysis of traffic.-In compiling tonnage figures, the port-toport traffic and the joint rail-water traffic were tabulated separately. The port-to-port traffic was tabulated according to type of carrier.

The barge-line revenue port-to-port (which does not include any terminal or transfer charges) for the common carrier and private or contract carrier was also compiled separately. These analyses of traffic are shown in the following table:

TABLE 22.-Analysis of traffic—Tombigbee route

It will be noted that about 78 percent of the traffic would be handled by private or contract carrier, 14 percent by common carrier, and 8 percent would be joint-haul traffic. These figures check very closely with the actual percentages of freight handled by contract and common carriers on existing improved waterways. From a comparison of the theoretical line-haul costs shown in paragraph 73, the bargeline revenues per ton-mile appear to be more than adequate. The common-carrier revenues are somewhat higher than Federal Barge Line revenues on the Mississippi and Warrior Rivers. The higher revenues reflected in this study are due to a large extent to the fact that many items of commerce were eliminated from consideration where barge-line revenue was considered not compensatory. An analysis of the Federal Barge Line's reports shows that on some items revenue accruing to the Federal Barge Line is as low as 1 mill per ton-mile. These low rates tend to bring down the average revenue per ton-mile. The same is true in general of private and contractcarrier operations.

81. Geographical distribution of commerce, Tombigbee route.-In the following table there is presented an analysis of the commerce on the proposed waterway, showing the distribution of the traffic according to various territorial subdivisions. These various subdivisions are shown on chart No. 6,' following page 59. This table indicates that movements of commerce are confined principally between points in the Tombigbee Valley itself, between the Gulf coast and the Tennessee Valley, and between the Gulf coast and the upper Tombigbee Valley. Less than 120,000 tons of commerce would move between the southern territory and other territories.

1 Not printed.

TABLE 23.-Geographical distribution of commerce, in tons

82. Density of traffic on proposed waterway.-The estimate of prospective commerce over the proposed waterway, as developed in the traffic survey, even with the arbitrary increase of 25 percent, appears to be an extremely conservative estimate when it is compared with the density of traffic moving on the Warrior River system in recent years. During the period 1928-32, there was an average annual movement of 1,586,000 tons of commerce on the Warrior River system. (This period includes both boom and depression years and is, therefore, believed to be fairly representative.) Dividing this average annual commerce by the 1930 population in the area tributary to this river system, 1,153,000, gives a movement of 1.4 tons per capita. Adding the existing tonnage on the Tennessee River, less the tonnage of Government material to the prospective tonnage for the proposed waterway and dividing the total by the estimated population in its tributary area in 1935, 2,824,320, the per capita movement for the proposed waterway is only 1.1 tons. This per capita tonnage appears to be low. In view of the potentialities of the tributary area, if the factor of 1.4 tons per capita of the Warrior River system be applied to the estimated population in 1935 in the tributary area of the proposed waterway, the total movement would be 3,954,000 tons.

83. Probable trend of commerce.-The figures contained in the tabulations of prospective commerce are a fair estimate of the commerce which would have moved in 1937. In order to measure with some degree of accuracy the probable traffic for the proposed waterway in the reasonably near future, a careful investigation of the more important factors which influence commerce in the tributary territory was made and a composite trend for all of the factors was developed. The trends of the following factors over a period of 29 years were

Total shipments