the availability of this remedy before the Commission which most strongly argues against court review of a permissive general revenue order. And this same rationale applied in the case of Commission action which vacated during the seven month statutory period-a suspension order with respect to rail, water or pipeline rates that had previously been imposed. See Amarillo-Border Express, Inc. v. United States, 138 F. Supp. 411 (N.D. Texas 1956), judgment vacated as moot, 352 U.S. 1028 (1957) (limited reviewability of the vacating order); accord, Long Island R.R. v. United States, 140 F. Supp. 823 (E.D.N.Y. 1956); Dixie Carriers, Inc. v. United States, 143 F. Supp. 844 (S.D. Texas 1956), judgment vacated as moot, 355 U.S. 179 (1957); Atlantic Coast Line R.R. v. United States, 173 F. Supp. 871 (E.D. Va. 1958). Contra, Freeport Sulphur Co. v. United States, 199 F. Supp. 913 (S.D.N.Y. 1961) (the vacating of a suspension order is reviewable only on grounds of abuse of discretion); Oscar Mayer & Co. v. United States, 268 F. Supp. 977 (W.D. Wis. 1967); Naph-Sol Refining Co. v. United States, 269 F. Supp. 530 (W.D. Mich. 1967). The orders in the cited cases were not the result of exhaustive hearings on the reasonableness of the proposed rates. More importantly, they-like general revenue orders-preserved the shipper's right to a full Commission hearing on the lawfulness of the new rates in a reparation proceeding. See, e.g., Naph-Sol Refining Co., supra.

In our case, by contrast, the motor carrier's proposed tariff schedule has virtually nothing to do with revenue requirements, general or otherwise. Rather, it seeks to make specific revisions in the charges for varying weight categories so that each might more properly bear its own share of the transportation burden. Moreover, the Commission's report and order here was not at all preliminary in character. It followed upon a most detailed investigation and specifically finds that the particular rates are reasonable both absolutely and in relation to one another.

The conclusive character of this report and order is emphasized by the fact that-in contrast to the statutory scheme with regard to rail carriers the Motor Carrier Act makes no provision for a reparation proceeding before the Commission. Hence, unlike a general revenue authorization for rail carriers or an order vacating a prior suspension of rail rates, the finding of "just and reasonable" contained in the report of October 28th is not readily susceptible of further challenge before the Commission itself. Of course, the shippers are not precluded from bringing a complaint sometime in the future under § 216(e), alleging that the Tariff 500 charges are unjust and unreasonable or at least will be so if applied from that time forward. However, such a proceeding would not reexamine the lawfulness of the rates at the time of the October 28, 1969, order. Hence, regardless of its outcome, it is unlikely under the Arizona Grocery doctrine that the findings in such a proceeding could serve as the basis for a judicial reparations action pursuant to § 204a (5) added to the Motor Carrier Act in 1965. Furthermore, as a practical matter, it is most improbable that the Commission-having

just completed a lengthy § 216 (g) investigation-would entertain a § 216(e) complaint or come up with significantly different findings in the immediate future. The report and order in this case are thus well within the ambit of reviewability recognized in Frozen Food Express v. United States, 351 U.S. 40 (1956), Abbott Laboratories v. Gardner, 387 U.S. 136 (1967), and Gardner v. Toilet Goods Ass'n, 387 U.S. 167 (1967). What relief may be appropriate is a different question, which we will discuss at the conclusion of this opinion.

III

The motor carriers sought to meet their burden of showing Tariff 500 to be just and reasonable, § 216 (g), 49 U.S.C. § 316 (g), primarily by an impressive revenue and expense study. Traffic data for 1967 were assembled for 42 carriers which accounted for 80% of the total revenue from traffic covered in Tariff 500. Sample freight bills were drawn by statistical procedures which the Commission permissibly found satisfactory.10 Revenues were then adjusted to 1968 levels. Costs for 1968 were divided into the five items mentioned in the margin 11 and assigned to the various weight groups under principles approved by the Commission and generally not attacked by the plaintiffs. Operating ratios were then derived for the various weight groups under the existing and proposed rates. Under one method, called Appendix B, the costs of the 20 group 1 carriers were applied as a system to the traffic of all 42 carriers studied; under another, called Appendix D, the costs of each carrier were applied to its own traffic. The Commission found the former method "less acceptable," 335 I.C.C. at 569. The Appendix D method yielded the following operating ratios, 335 I.C.C. at 585:

10 The shippers assailed the sampling technique employed by the carriers on the ground that the selection of freight bills was drawn on a systematic (e.g., very fourth or every eighth invoice) rather than a random basis. Such a procedure will not produce a biased sample, however, unless there are periodicities in the data, and the carriers' sampling experts demonstrated through various randomness tests that there were no marked periodicities in the sample data. 335 I.C.C. at 566.

11 Pickup & Delivery at Origin and Destination; Pickup & Delivery at Interchange; Platform; Billing & Collecting; and Line Haul.

This indicated that traffic in the 101-499 pound range was being carried at a loss under existing rates, whereas Tariff 500 would produce reasonable operating ratios, and constituted the principal basis for the Commission's conclusion that the carriers had sustained their burden of justifying higher charges for small shipments.

Plaintiffs mount two major attacks upon the Commission's approval of the carriers' cost studies. One relates to the treatment of so-called constant costs, the other to the inclusion of some non-study carriers' traffic which was handled interline with the study carriers. We reject the former but sustain the latter.

The Commission's Cost Section has long held the view, not here attacked, that 90% of the costs of motor carriers vary with the amount of traffic ("out-of-pocket" costs), whereas 10% are 'constant. 12 Before the agency the shippers asserted that only out-of-pocket costs should be considered a position not seriously pressed here and which the Commission properly rejected on the ground that in a proceeding concerned with the totality of traffic, no cost element could be properly neglected. They argued alternatively that constant costs should be allocated to the various weight groups by first dividing these between terminal costs (which were subdivided into billing & collecting, pickup & delivery, and platform) and line-haul costs on the respective ratios of such costs, and then assigning the former to shipments on the basis of total hundred weights and the latter on the basis of total hundred weight miles—a procedure sanctioned by the Commission's Cost Section for some purposes. The carriers urged that however appropiate this method might be in other types of proceedings, in a case whose aim was a proper distribution of the entire cost burden constant costs should be allocated as a percentage addittion to the out-of-pocket costs determined by the studies for each weight bracket. The Commission upheld this position, 335 I.C.C. at 573-75.

While neither method of allocation is perfect, the Commission had a rational basis for sanctioning the carriers' decision to distribute con

12 These would include, for example, a portion of the costs for maintenance of plant and equipment, that part of depreciation which is not wholly a function of use, and "general expenses" such as the salaries of higher officials and the expenses of maintaining the general offices of the company. Cf. Locklin, Economics of Transportation 131-34 (6th ed. 1966). Unlike the railroads, however, the motor carriers are not shouldered with the direct responsibility of maintaining their own roads, which constitutes a significant element in the constant costs of the former. Id. at 132.

stant costs by dollars rather than by physical units. As it pointed out, application of protestants' method would create absurd disparities in the additions to out-of-pocket costs for different weight groups, as shown in the margin.18 We know of nothing in the nature of things or in recognized economic or accounting principle which dictates that those costs which comprise the 10% regarded as "constant" must inevitably be distributed on the basis of physical units rather than on the basis of expenses directly attributable to handling them. While plaintiffs may be right in saying that the Commision's determination sanctions a rate structure which more closely resembles one based on the assumption that all costs are variable-a position the carriers defended and the Commission rejected, 335 I.C.C. at 573 than it would were the cwt./ cwt.-mile allocation employed, we see no reason to overturn the determination on that score. Indeed plaintiffs' cost analyst conceded that "[i]n the long run, over a twenty-five year period, all costs are outof-pocket, all costs are a hundred percent related [to] traffic if we went a long enough period of time." And it has been said that "[i]n the motor-carrier industry. . . the long run is short in terms of calendar time, unlike the situation in industries which have difficulty in adjusting their plant and equipment to changes in the volume of output. Locklin, Economics of Transportation 648 (6th ed. 1966). In any event, a matter of this sort is peculiarly for the Commission to determine.

The other attack on the cost study is much more serious. The study included three types of traffic: (1) traffic which each of the 42 study carriers transported from origin to destination on its own line; (2) traffic which one study carrier originated and turned over to another carrier whether a study carrier or not, for transportation to destination; and (3) the portion of non-study carriers' traffic which was received from or delivered to a study carrier. A cost analysis based on all three types of traffic was deemed to be made on a "through basis'' since it included the complete haul on all traffic in which study carriers participated. This is to be distinguished from the "carried basis" which was limited to the study carriers' own haul, i.e., the first two categories noted above. Plaintiffs object that since interline traffic involves nonline-haul costs additional to those of single line traffic, apparently of the order of a 4-5% increase in operating ratio, inclusion of interline traffic-and the attendant interchange costs of non-study carriers without including their single line traffic produces a general upward bias in the cost sample and a particular distortion with respect to small shipments, which are appropriately charged with a much higher proportion of non-line-haul than of line-haul costs. The proper method would thus

13 Percentage Increases in Out-of-Pocket Cost by Allocating Constant Costs on Hundred Weight-Hundred Weight Mile Basis.

have been either to include the single line traffic of the non-study carriers as well as their share of traffic handled interline with the study carriers or, since that was evidently impracticable, to confine the study to the traffic handled by the study carriers on their lines, the "carried basis.'' 14

The Commission makes no real attempt to justify its approval of the "through basis," and ECMCA's defense of it hinges on a misstatement of the plaintiffs' claim. The main argument of both defendants is that the error was harmless. The Commission presents a calculation from exhibits of record, on a "carried basis," showing the following operating ratios:

Since these figures also show unreasonably high operating ratios in the 101-499 pound bracket under the existing rates although lower than those on which the Commission relied, and reasonable ones under the proposed charges, although again lower than those cited in the report, counsel for the Commission urges us to disregard its error as inconsequential.

We would have some trepidation in affirming an agency's decision on the basis of complex calculations made on brief although from exhibits of record, where its opponents have been afforded no opportunity to cross-examine or to rebut the calculations by testimony, see Davis & Randall, Inc. v. United States, 219 F. Supp. 673, 682 (W.D. N.Y. 1963), even though they have done this on brief and at argument to good

14 We reject ECMCA's contention that this point is not properly before us because "[p]laintiff neither raised the argument nor submitted evidence in support of it during the course of the hearing." The National Small Shipments Conference clearly raised the point in its brief to the Commission, and the agency passed upon it, 335 I.C.C. at 570.