mately 300 and competition with rail carriers to Washington and other points was intensified.

The rail carriers are at some disadvantage in competing with truck carriers for the transportation of automobiles, in that, generally speaking, the truck carrier offers a more elastic and speedier service than the rail carrier, and the shipper is out to an additional expense of approximately $2 per automobile in removing bumpers and otherwise preparing machines for shipment in box cars. Interveners point out that before the 50-cent rate was published the rate for transportation by truck was no lower than the rail rate and they urge that the reduction of 12 cents in the rail rate was an effort to destroy motor transportation. Interveners emphasize the failure of applicants to show the cost of performing the service, and endeavor themselves to approximate the cost as hereafter explained. The record contains no evidence as to the value of railroad cars equipped with Evans loaders or kindred appliances and used for the transportation of this traffic, or as to the extent of the empty movement. As shown at the first hearing, the average loading from April 15 to May 10, 1935, was 11,575 pounds. The distance from Detroit to Chicago over the Michigan Central-New York Central route is 281 miles, and the average revenue was $57.88 per car, 20.6 cents per car-mile, and 3.5 cents per ton-mile. This distance is only 9 miles greater than the short-line distance. The average loading for June 1937 was 11,897 pounds, resulting in revenues of $59.49 per car and 21.2 cents per carmile for a haul of 281 miles. The distance over the most circuitous route from Detroit to Chicago is 396 miles. For this distance the average revenue of $59.49 per car yields 15 cents per car-mile. The average car-mile revenue of the New York Central on all traffic in 1936 was 22.6 cents and the average haul 217 miles. Assuming the same tare weight of equipment for all traffic as for automobiles, for each net ton of revenue-paying freight of all traffic the New York Central hauled an average gross weight of 2.171 tons, while on the automobile traffic it hauled an average gross weight of 5.87 tons. Of the average gross weight of loaded cars of automobiles only 17 percent was revenue-paying weight, while, of all freight, 47 percent of the average gross weight was revenue-paying weight. For each ton of automobile revenue traffic the carrier transported an average of 4.875 tons of car, and for each ton of revenue freight of all traffic an average of 1.171 tons. However, the yield per car-mile was not materially lower than the average for all traffic.

Interveners contend that automobiles should pay substantially greater revenue than a carload of average freight, because of the nature of the traffic, the high class of equipment required, the expense

of store-door delivery not usually given on carload traffic, and the amount of tare weight for each ton of revenue freight. That contention is sound as respects a maximum reasonable rate on automobiles. It is, however, the privilege of applicants to maintain a rate lower than a maximum reasonable rate if it be no lower than the minimum of reasonableness and no lower than is necessary to meet the competition of the motor carriers.

The Plymouth and Dodge plants at Detroit are on the New York Central, and the Chrysler, De Soto, and Dodge truck plants are on the Detroit Terminal Railroad. Between April 15, 1935, and June 30, 1937, the Michigan Central absorbed switching of $7 per car on 3,844 carloads of this traffic. Interveners assume that there was a per-diem charge on these cars of $3 per car as they were loaded on a connecting line, making a total charge of $10 per carload. As 19,949 carloads were handled by the Michigan Central during this period, the total of $38,440 would average $1.93 on all cars handled. Deducting the latter amount, plus $9.85, the cost of store-door delivery in Chicago, from the average car revenue of $57.88 above mentioned would leave $46.10 for furnishing and placing the empty car, hauling the load to Chicago, and returning the empty to Detroit. For the haul of 281 miles this is equivalent to 16.44 cents, and for the haul of 396 miles 11.64 cents, per loaded-car mile. But of course a part of the average cost of furnishing and placing empty cars was embraced in the deduction made of $1.93 per car. Furthermore, the June 1937 average loading would produce car revenue of $59.49.

Interveners point out that in Switching Rates in Chicago Switching District, 177 I. C. C. 669, 195 I. C. C. 89, we approved a minimum switching rate of $18 per car, and contend that the service performed in moving this traffic from the break-up yard to the unloading dock at Twenty-third Street, and returning the empty car, is in all respects similar to an ordinary one-line switching service performed at Chicago. In Increased Switching Charges at Detroit, Mich., 91 I. C. C. 82, division 4 found that the Michigan Central and New York Central had justified charges of $16 per car for intraterminal switching, $12 for interterminal switching, and $7 for reciprocal switching, and in Switching at Detroit, 95 I. C. C. 21, the division found justified switching charges of $16 and $12 for intraterminal and interterminal switching by other principal railroads at Detroit. Extensive cost studies were presented in all three of those proceedings, and the carriers' proposed charges found justified, at least at Detroit, are less than the cost of rendering the service as developed by those studies. For instance, respondents' cost study in the first Detroit proceeding developed costs for interterminal and reciprocal switching to and from industries of $15.26. The division's comments on

the cost studies appear at pages 93-96 of the report. Studies presented in the second Detroit proceeding developed average switching costs of $15.41, $13.63, and $15.82. If the switching charges approved in those cases are at all indicative of the cost of movements within those terminals, say interveners, there is little left out of the per-car revenue under the 50-cent rate for the furnishing of the car and the road-haul movement. Thus $12 for the switching at Detroit, $18 at Chicago, and $9.85 for store-door delivery totals $39.85, leaving, out of $57.88 per car, but $18.03 for furnishing and upkeep of cars equipped with Evans loaders, road haul of the loaded cars, and such movement of the empty cars as may be required, which return, it is urged, is obviously not reasonably compensatory.

Considering this traffic to be added traffic that will move by rail only at low rates and assuming that the present facilities of the rail carriers are not used to capacity, the cost that may be allocated thereto is the additional cost above the cost of operations without this traffic. Such added cost is much less than the average cost of transporting all traffic. Absorptions of switching and costs of storedoor delivery are of course added costs to be fully allocated.

Applicants maintain that the 50-cent rate is reasonably compensatory, and in rebuttal they rely upon the comparison of the car-mile revenue thereunder with the average car-mile revenue on all traffic handled by the New York Central, hereinabove set forth, and also a comparison of the ratio of these revenues with the ratio of the firstclass rates for the respective distances. The first-class rate in central territory for 217 miles, the average haul hereinbefore given, is 76 cents, and for 281 miles 87 cents, and the ratio of these rates per mile is as 0.35 cent to 0.31 cent. The ratio of the car-mile revenue on all freight to that on automobiles is as 22.6 cents to 21.2 cents. A ratio of the car-mile revenue on all freight to a car-mile revenue of 20 cents would be the same as that of the first-class rates for the distances stated.

Applicants also refer to Automobiles and Parts to Southern Territory, 215 I. C. C. 488, Automobiles to Oklahoma City-Ada-Atoka Ry. Points, 219 I. C. C. 314, Automobile Chassis to North Carolina Points, 220 I. C. C. 201, and Automobiles to Arkansas and Louisiana, 220 I. C. C. 794, wherein fourth-section relief was granted in connection with rates on automobiles and chassis. A rate of 66 cents found reasonably compensatory in the first of those proceedings which over a circuitous route of 332 miles from Cincinnati, Ohio to Jellico, Tenn., produced car-mile revenue of 19.9 cents at the carload minimum of 10,000 pounds for the smallest size of car. That rate over the direct route of 214 miles produced 31.3 cents per car-mile. Similarly compared on the same minimum-weight basis,

the 50-cent rate over the most circuitous route from Detroit, 396 miles, produces 12.6 cents per car-mile, and over the route of movement 17.6 cents. In the second of the proceedings mentioned, a rate of 69 cents was found reasonably compensatory which over a circuitous route of 484 miles, 29 percent circuitous, from Kansas City, Mo., to Tupelo, Okla., yielded 16.8 cents per car-mile at the average loading for passenger vehicles of 11,800 pounds. The third proceeding mentioned covered chassis only. The rates therein found reasonably compensatory over the circuitous routes, at the minimum of 10,000 pounds, were 55 cents for 422 miles from Norfolk, Va., to High Point, N. C., yielding 13 cents per car-mile and 85 cents for 505 miles from Baltimore, Md., to Thomasville, N. C., yielding 17 cents per car-mile. In the fourth proceeding cited the report states that the proposed rates, at the average loading of 11,500 pounds, will yield car-mile revenue ranging from 17.21 to 23.11 cents. The illustrative rates given in the report, and the car-mile yields thereunder, are $1.86 from Evansville, Ind., to Alexandria, La., circuitous route 938 miles, yield 22.8 cents, direct route 649 miles, yield 32.9 cents; and $1.26 from Evansville to El Dorado, Ark., circuitous route 805 miles, yield 18 cents, direct route 540 miles, yield 26.8 cents. None of the rates approved in those proceedings included store-door delivery.

Applicants refer also to low car-mile yields of rates approved by us for relatively heavy-loading commodities such as burlap bagging, grain and grain products, and molasses in tank cars, supporting their contention that the net car-mile yield of the 50-cent rate is compensatory.

The fourth-section departures brought about by the 50-cent rate arise from the fact that this rate if used in combination with rates from Chicago to points in Illinois, Wisconsin, Minnesota, Iowa, and Missouri, would result in rates ranging from 1 to 12 cents lower than the published through rates. In the prior report division 2 quoted from Commodity Rates in Official Territory, 209 I. C. C. 702, 705, the statement that:

Truck competition does not justify departures from the aggregate-of-intermediates provision to establish reduced rates at an intermediate point, if the through traffic is subject to competitive influences to the same extent as is that from or to the intermediate point.

The division found as a fact that the movement of automobiles to Chicago from the origins here considered is subject to more intensive truck competition than affects the movement of like traffic from the same origins to destinations beyond Chicago. Interveners contend upon the new evidence that the competition is not more intense at Chicago than at points west thereof.

From Detroit to many Missouri points and from both Flint and Detroit to many Iowa points and in other instances, the combination rates are less than the truck rates. To St. Louis, Mo., the largest consuming point in affected territory, from Detroit, the present rate is 4 cents higher, and the combination would be 2 cents higher than the truck rate, while from Flint the present rate is 5 cents higher than, and the combination would be the same as, the truck rate. The truck rates to Chicago from these origins are published in amounts per automobile, but at the shipping weights of automobiles they are generally the equivalent of 50 cents per 100 pounds, or very close thereto.

Three trucking companies transported from Detroit and Flint to Illinois points, other than Chicago, and to points in Iowa and Missouri, 12,648 automobiles in 1935, 12,141 in 1936, and 10,032 during the first six months of 1937. The only statistics of record of the movement by rail to affected territory are for the Michigan Central, covering the first three months of 1935 and the first four months of 1937. The movement to Illinois points other than Chicago and to points in Iowa and Missouri for the three months of 1935 was 279 carloads from Detroit and 54 carloads from Lansing. Assuming four automobiles to the carload, and projecting for the full year by multiplying by 4, would make a rail movement of 5,328 automobiles, less than half of the truck movement. The movement of the four months of 1937 was 199 carloads from Detroit and 77 carloads from Lansing. Projecting in the same manner for a six months' period would make 1,656 automobiles, as contrasted with 10,032 moved by truck. All shipments to St. Louis from these origins move by truck. The evidence indicates that there is no substantial volume of movement by truck from these origins to Minnesota and Wisconsin. To Chicago four trucking companies transported from Detroit, Pontiac, and Flint during 1935, 1936, and six months of 1937, 65,843 automobiles, as compared with 19,949 carloads, or 79,796 automobiles, assuming four to a carload, transported by the Michigan Central from April 15, 1935, to June 30, 1937. The latter is the only rail carrier which is now receiving more than an occasional carload under the 50-cent rate.

So far as these statistics go, they indicate, except in Minnesota and Wisconsin, just as keen competition between rail and truck carriers at many points in the territory covered thereby beyond Chicago, as at Chicago. The statistics are, however, incomplete. In the affected territory, they cover the traffic of only one rail carrier and only three trucking concerns. Approximately 45 companies engage in trucking automobiles from the origins here considered to Chicago or points beyond.