(dissenting).
This is the second time this panel has considered this appeal. When it was first before us, the majority filed an opinion reversing the district court, charging that court with making a finding which was clearly erroneous. I filed a dissenting opinion. Following a petition for rehearing, there was a reconsideration by the same panel. The majority now find as clearly erroneous that which they specifically found in their first opinion as “not clearly erroneous.” I again dissent. This time I dissent not only from the conclusion reached by the majority but also the method utilized to reach it.
This appeal requires review of a non-jury determination that a shipper failed to establish a prima facie case that shipments of hams were in good condition at the time of delivery to a railroad carrier. At trial, the shipper relied upon an acknowledgment in the railroad’s bill of lading that the goods were “in apparent good order.” The railroad contended that this acknowledgment in the bill of lading fails to satisfy the prima facie evidence rule because the phrase refers only to shipments “which were open to inspection and visible,” Tuschman v. Pennsylvania Railroad Co., 230 F.2d 787, 791 (3d Cir. 1956), whereas here the hams were received by the railroad in sealed trailers.
The district court found that the goods were loaded into trailers by the shippers, and that these trailers were then sealed and placed on the railroad’s freight cars for transportation. Based on its finding that the trailers were sealed, the court concluded that the railroad was not bound by the “apparent good order” statement on the bill of lading because “[w]here merchandise is sealed in a trailer . . . the contents are not open and visible.” 329 F.Supp. 1118.
The majority determine that application of the clearly erroneous rule dictates rejection of this critical finding. When this appeal was before this panel for original consideration, the majority stated that “the district court’s finding that the trailers were ‘sealed’ by the shipper before delivery to the carrier is supported by substantial evidence and is not clearly erroneous.” The majority properly observed in their first opinion that the “later stipulation of the parties concerning Plan 2-¼ did not by its terms negate the testimony of Mr. Roeder.” The majority now hold there is not “a sufficient evidentiary basis to support a finding that these trailers were sealed.” (page 107.).
The majority attempt to justify this oscillation by ascribing some significance to certain statements made by appellate counsel, not at trial in the district court, but adduced at a pre-argument hearing conducted after the original decision by this court, thirteen months after the district court trial took place. I am in total disagreement with this approach and would affirm the judgment of the district court.
What divides this court, after reconsideration, is basically the same difference in jurisprudential philosophy which divided it upon initial consideration. Ours is a fundamental difference in what we perceive to be the appropriate role of an appellate court in reviewing facts found by a trial court sitting without a jury. Regretfully, there is very *109little in the approach taken by the majority with which I can generate even a modicum of agreement.
The district court observed:
Critical to our decision are these facts. The rail carrier furnishes a piggyback trailer to the shipper, who cools and loads the shipment in a trailer which he is required to pre-cool. The shipper then seals the trailer. The trailer is then delivered to the piggyback train for cross country transit.
329 F.Supp. at 1117-1118.
The majority view, as clearly erroneous, the finding that the trailers were sealed. This was an important finding, for it was on this basis that the district court was able to draw the distinction “between those cases in which the merchandise being shipped were open to inspection and visible. Tuschman v. Pennsylvania Railroad Co., [230 F.2d 787, 3d Cir. 1956], and a situation wherein a loaded and sealed trailer is delivered to a carrier. Lincoln Farm Products Corp. v. Central Railroad Co. of New Jersey [81 N.J.Super. 161, 195 A.2d 200 (1963)].” 329 F.Supp. at 1118.
I find sufficient evidence in the record to sustain the trial court’s finding that the trailers were sealed. As the majority concede: (1) the shipments were made under “Plan 2-44” (page 104); and (2) it was the uncontradicted testimony of Mr. Roeder, a witness for the railroad: “In a plan two and a quarter, the shipper would affix the seal to the rear door, that trailer would be received at the ramp, or the trailer received under seal.”
Notwithstanding the uncontradicted record before the district court, the majority would by judicial fiat distort the trial record and infect it with dubious matters adduced at a F.R.A.P. 33 pre-hearing conference before an appellate court. With all due respects, I cannot detect the faintest tint of the relevancy of a post-trial, post-appellate argument disagreement “between counsel on the facts concerning the sealing of these trailers.” The function of an appellate court is to review only the record which was before the district court in order to determine whether that court committed error. Thus, I consider it highly improper for an appellate court to consider any evidence, stipulation, or statements of counsel which were not before the trial court at the time the decision under review was made. F.R.A.P. 10. Hunt v. Local Board No. 197, 438 F.2d 1128, 1146 (3d Cir. 1971), (dissenting opinion, Aldisert, J., joined by Van Dusen, J.).
Yet, that is precisely what the majority has done in reaching its conclusion. In an attempt to fashion some basis for the conclusion that there was insufficient evidence at trial “to support a finding that these trailers were sealed” (page 107), the majority rely on matters presented to an appellate court which were not before the district court. They rely on “comments of counsel” and “the Stipulation and statements thereafter filed [which] make clear there is no agreement between counsel on the facts concerning the sealing of these trailers and the existence or non-existence of a ‘custom in the industry for the party loading the shipment into the trailer to place its seal upon the trailer,’ .” (pages 106, 107.) Moreover, even if such a procedure did not offend my concepts of appellate review, the substance of post-trial disagreement of facts by trial counsel is totally inconsequential.
In concluding “there was not a sufficient evidentiary basis to support a finding that these trailers were sealed,” (page 107) the majority pointedly ignore Mr. Boeder’s testimony:
Q. Do you ever break the seals on those trailei-s?
A. Once in a while we do break the seals.
Q. For what purpose would you break the seals?
A. [A description of the circumstances.]
* * *
*110Q. Do you inspect all the loads that come into Philadelphia ?
A. No, we don’t.
Q. Where would that notation be if the seal were broken ?
A. We would make a notation on the waybill, and we would also make a notation in our seal record book.
Q. If you know, were the seals broken in any of these four cases?
A. Not to my knowledge.
(Record at 132-133.) (Emphasis supplied.)
The irrefutable and controlling fact is, as the majority explicitly conceded in their original opinion, that no evidence was introduced at trial to offset the critical evidence that the trailers were shipped under Rule 2*4, and Mr. Roeder’s testimony that the trailers shipped under Rule 2*4 were sealed.
Accordingly, I would affirm the judgment of the district court.