The S. S. Marie Loenhardt collided with the Delair Bridge at about 12:45 p. m. on January 9, 1959. The Pennsylvania Railroad Company, owner and operator of the bridge, filed a libel in rem against the vessel for damages to the bridge.1 The vessel owner filed a cross-libel against the Railroad for damages to the vessel resulting from the collision. The district court, after making findings of fact and conclusions of law, dismissed the Railroad’s libel but allowed the owner of the vessel to recover against the Railroad for whatever damages the vessel sustained. 202 F.Supp. 368 (D.C.E.D.Pa., 1962). From the interlocutory decree, the Railroad has appealed under 28 U.S.C. § 1292(a) (3).

The Delair Bridge2 is a low-slung, truss-type steel bridge that carries two sets of railroad tracks across the Delaware River between Philadelphia, Pennsylvania, and Delair, New Jersey. At the place of crossing, the river is 2,100 feet wide. Two-thirds across the bridge from the Pennsylvania side is a section or drawspan 300 feet long, equal to the width of the river channel. This section has a vertical clearance of 48 to 55 feet from the water, depending on the tide, and pivots 90 degrees counterclockwise upon a pier located at its center. The pier is in the middle of the channel. When the section is completely opened, two lanes are made available for the use of vessels and crafts proceeding up or down the river channel. Although the machinery which rotates tho drawspan is located on the central pier, the drawbridge tender stationed at the section can open the draw only after the operator in the Jersey Block Signal Control Tower has electrically released the lock. The Jersey Tower, a three-story building, is located inland about 750 feet east of the New Jersey end of the bridge. A radiophone is located in the Jersey Tower, but not on any part of the bridge. Communication between the Jersey Tower and the control room of the drawspan is via a telephone line.

Six hundred feet from the Pennsylvania shore, and starting at a line drawn across the river approximately 2,000 feet below the northeastern tip of Petty Island, the river channel narrows to a width of 300 feet. From this line, the channel runs diagonally in a straight line, northeastwardly across the river for about 5,000 feet to within 300 to 400 feet of the New Jersey shore. This stretch of the channel is called the Fisher Point Range. At the end of the range the channel bends slightly to the left and extends for another 1,800 feet in a straight line. At the end of this stretch, known as Fisher Channel, it bends again slightly to the left and continues on in a straight line for about 800 feet where it is crossed at right angles by the draw-span of the Delair Bridge. The stretch of the channel leading to and under the bridge is called the Draw Channel. Anchorage areas exist on both sides of the Fisher Point Range and on the Pennsylvania side of the Fisher Channel.3 *264Except possibly for a small area where the Fisher Point Range joins the Fisher Channel, the river on the New Jersey side of the channel described above is too shallow to accommodate large vessels with drafts over twenty-three feet. The anchoring of a vessel is not permitted on either side of the Draw Channel even though the river on the Pennsylvania side is deep enough to float larger vessels.

The S. S. Marie Leonhardt, an oceangoing vessel of 15,000 tons (deadweight), carrying iron ore, was enroute up the Delaware River from Pier 122 South, Philadelphia, to Morrisville, Pa., accompanied by the tug Joan McAllister. She was 522 feet long, had a beam of over 63 feet. She was on an even keel with a draft of 23 y2 feet and was propelled by a single right-hand screw. In order to reverse the screw, her diesel engines had to be stopped and then started again in the other direction. She was equipped with a portable radio-phone set having a range of at least seven miles. She was unable to pass under the closed drawbridge of the Delair Bridge. During the vessel’s passage up the river before she came into contact with the bridge, the tide was at flood with a two-knot current, there was a steady wind from the northwest at 22 knots per hour, with gusts up to 27 knots, and the air was clear and visibility was excellent. There were no vessels or crafts coming downstream on the other side of the bridge desiring to pass through the draw.

At the outset the Railroad disputes the findings of the trial judge. Of course, such findings will not be disturbed on appeal unless they are clearly erroneous. McAllister v. United States, 348 U.S. 19, 75 S.Ct. 6, 99 L.Ed. 20 (1954); Brett v. J. M. Carras, Inc., 203 F.2d 451 (C.A. 3, 1953); The Bellatrix, 114 F.2d 1004 (C.A. 3, 1941). In the alternative, it insists that the facts, as found, establish negligence on the part of the vessel and no negligence on its part.

Citing Patterson Oil Terminals, Inc., v. The Port Covington, 109 F.Supp. 953, 954 (E.D.Pa. 1952), affirmed 205 F.2d 694 (C.A. 3, 1953) and Wilmington Ry. Bridge Co. v. Franco-Ottoman Shipping Co., 259 F. 166 (C.A. 4, 1919), among other cases, the Railroad contends that the district court should have given it the benefit of the presumption that the vessel was negligent because she collided with a fixed portion of the bridge. We are hard pressed to understand why the Railroad is making this contention at this point in the proceeding. Perhaps it is implying that the presumption is a makeweight in the evidence which would require the mythical scales on which conflicting testimony is weighed to be tipped in its favor. If this is so, then the Railroad misconceives the function of the presumption. The vessel owners complied with the procedural requirement of the presumption. As a matter of fact, both sides fully presented testimony regarding their version as to what happened prior to the collision. Consequently, the presumption disappeared as a rule of law. See 9 Wigmore on Evidence (3rd Ed.) §§ 2490, 2491.

The Railroad next claims that the vessel brought itself within the rule of The Pennsylvania, 86 U.S. (19 Wall.) 125, 136, 22 L.Ed. 148 (1873), by approaching the Delair Bridge both without complying with the Bridge Regulations of the United States Army Corps of Engineers regarding that bridge, and on the wrong side of the channel in violation of the narrow channel rule, and has failed to show that these violations could not have been a cause of the collision. We disagree.

*265 The Railroad says that the vessel violated the regulations 4 by choosing to initiate communications with the bridge by radiophone instead of by whistle or horn signal. The district court found that the vessel, in conformity with the regulations, sounded three blasts of her whistle at 12:40 to indicate that she desired to pass through the draw. The bridge did not answer this signal by any whistle or horn blast.5 The drawbridge tender testified that he got a signal from the Jersey Tower operator at 12:36 that the draw section was free to open. If this were so, then he should have answered the vessel’s three blast whistle signal at 12:40 with a two blast signal. He claims that he did not do so because he did not hear the vessel’s three blast signal. If the wind prevented the bridge tender from hearing a whistle signal given a half mile away, it would havC' prevented him from hearing one given at'any time between 12:26 and 12:40 whem the vessel was further downstream, and! so it would have been futile for the vessel to have given a whistle signal sooner'. Hence the Railroad cannot be heard to complain that the vessel did not give a whistle signal at any time before 12:40.

Concerning the narrow channel rule,6 the district court held it was inapplicable because it “is designed to prevent collisions between vessels on inland waterways, rather than collisions involving a ship and a bridge.” 7 The rule was invoked by the owner of a bridge in Circle Line Sightseeing Yachts, Inc. v. City of New York, 283 F.2d 811, 814 (C.A.2, 1960). Although the Court of Appeals ruled that adherence to the narrow chan*266nel rule was not “safe and practicable” because of the peculiar and quite temporary circumstances existing there, it went on to hold, nonetheless, that the vessel was at fault because of her failure to show that her violation of Article 29 of the Inland Rules, 33 U.S.C.A. § 221,8 did not cause and could not have been the cause of the accident. Although another vessel proceeding in the opposite direction was involved there, and while that case is good authority for holding that the owner of the bridge may bring the rule of The Pennsylvania, supra, into play where a vessel violates the Inland Rules, nevertheless, it must be shown that that violation could not have caused the accident, as adverted to above. Here, the Marie Leonhardt’s proceeding in the middle or to the port (Pennsylvania) side of the channel was a mere condition and could not have been a contributing or substantial cause of the collision. It could not have been any more so than was the vessel’s departure from Pier 122 South, Philadelphia, shortly after 11:30, instead of at a later time. Moreover, it was practicable for the vessel to favor that side of the channel, at that time in the absence of downstream traffic. The river on the Pennsylvania side of the channel in that vicinity was deep enough to accommodate vessels with drafts of 23y2 feet, and a moderately strong wind was on the port side of the vessel, and she was proceeding at the slowest possible speed under the circumstances. For these reasons the pilot wanted to be as near to the anchorage area as possible so that he could get out of the channel in the least amount of time in the event the vessel was required to drop anchor.

The Railroad next argues that the vessel was at fault in proceeding at too great a speed and too close to the bridge in the absence of a horn or whistle signal from the bridge. This argument has been adequately answered by the district court. Its findings as to the happenings between 12:26 and 12:48 on January 9, 1959, may be summarized as follows:

12:26: The vessel was in the vicinity of Port Richmond approximately 11,200 feet from the bridge. In accordance with the prevailing custom, the pilot of the vessel first informed the bridge by radiophone of the vessel’s presence and of his intention of sailing her through the draw. Immediately thereafter, the Jersey Tower operator notified the drawbridge tender of the vessel’s intention.

12:34: The vessel was about a mile and a fifth away from the bridge (about half way up the Fisher Point Range), and the pilot again communicated by radiophone with the tower operator who advised him that an approaching train would require a two or three minute delay in opening the draw. The pilot made no reply to this response. From that time the vessel moved at the slowest possible speed, estimated at 4 to 6 knots per hour over the ground, in order to maintain steerageway in the channel. She favored the Pennsylvania side of the channel.

12:36: A two car train, some four or five minutes behind schedule, cleared the movable span, and the vessel was slightly over two-thirds of a mile away (or slightly over the three-quarters up the Fisher Point Range).

12:39: The pilot, having observed the passage of the train over the span, again spoke to the tower operator by radiophone. The vessel was about 3,000 feet away (or a few hundred feet from the Fisher Channel) and moving steadily nearer. The tower operator responded that the bridge was opening, although at this time no movement of the drawspan was observed by the vessel. Just prior to 12:40 the vessel blew three blasts on its whistle; no whistle or horn signal was sounded by the bridge in response.

12:40: The vessel was approximately one-half mile from the bridge (or entering the Fisher Channel). In view of the *267tower operator’s responses over the radiophone at 12:34 and 12:39, and in the absence of a horn or whistle response to the vessel’s three blast whistle signal or any signal indicating further delay in opening the bridge, the pilot abandoned any plan of dropping anchor in the area on the Pennsylvania side of the channel. The vessel continued to proceed up the channel. Immediately prior to 12:42, the vessel was in a reasonably good position and under sufficient control to pass successfully through the draw had it been opened.

12:42: Confronted with the unopened bridge when the vessel was 900 feet away from the bridge (or one hundred feet from the Draw Channel), the pilot ordered the dropping of the starboard bow anchor and the engines full astern. The tug Joan McAllister was summoned to push on the vessel’s port bow. Simultaneously with the dropping of the anchor, the drawspan began to open.

12:43%: The draw was fully opened, 17% minutes after the vessel first communicated with the tower.

12:45: The bow of the vessel rammed through the wooden fender rack surrounding the pier of the first fixed section of the bridge to the west of the drawspan and became wedged therein.

12:48: The port quarter of the vessel settled against the fixed steel structure immediately west of the drawspan.

When the pilot of the vessel received word from the tower operater at 12:34 he had two choices: Either turn to port and drop anchor in the area immediately adjacent to the Pennsylvania edge of the channel, or time the opening of the draw-span by moving slowly up the channel and arrive at the bridge slightly after 12:40 when he expected the draw would be opened. Large vessels cannot stop immediately or hover in one spot in narrow channels by turning around in circles; they must either drop anchor or proceed above a minimum speed in order to maintain headway. The Railroad says that the vessel should have proceeded to the anchorage area because the bridge did not give the vessel any whistle or horn signal.

The regulations are silent as to what the vessel should do in the event the bridge does not give a responsive signal. But the regulations govern only a part of the duty owed the bridge by the vessel. As pointed out in Shell Petroleum Company v. Peschken, 290 F.2d 685, 687 (C.A. 3, 1961), there is a line of cases “where it has been said that in the absence of warning to the contrary a ship approaching a bridge may assume the bridge will be open v/hen it arrives.” We need not rely on those cases. This is not a situation where a vessel proceeded toward a closed drawbridge just because she had given a signal requesting passage through the draw. There was more than the absence of a responding signal. The district court found that the conduct of those in charge of the bridge constituted an invitation to the vessel to proceed, that the pilot of the vessel reasonably inferred, in reliance on that invitation, that the drawspan would be timely opened and permit the vessel’s passage, and that the “failure of the drawtender to respond to the whistle signal cannot be held to negate the strong invitation previously extended.”9 These findings are not clearly wrong.

Concerning the cross-libel, the Railroad claims that it was not at fault. While the district court held that those in charge of the bridge did not violate the bridge regulations, yet the Railroad cannot successfully argue there was no duty cast upon it to act affirmatively until the responsible employees received a three blast signal from the vessel. The testimony shows that the operator and the drawbridge tender were unfamiliar with the provisions of the bridge regulations and further that, had they heard a three blast signal, they would have responded with a one blast signal, indicating they had heard the three blast signal. Yet, if the vessel had repeated the three blast signal after 12:40, until it was per*268ceived by those in charge of the bridge, the employees would have answered with a one blast signal, which would have been -the wrong signal.10 The regulations, as in the case of the duty owed the bridge by the vessel, does not encompass the bridge’s entire duty owed vessels desiring to pass through the draw. There is no . statute or regulation expressly requiring ■ those in charge of the Delair Bridge to . adopt radiophones as a means of communications. But the absence of such a ■requirement does not absolve them from ■the duty of due care in using that means 'of communication once they have voluntarily adopted it. See Indian Towing Co., Inc. v. United States, 350 U.S. 61, 76 S.Ct. 122, 100 L.Ed. 48 (1955); Erie R. Co. v. Stewart, 40 F.2d 855 (C.A.6, 1930). It was by means of this advanced method of communication that the tower operator gave information to the vessel. Part of that information was misleading. Both the tower operator and the drawbridge tender knew as early as 12:26 p. m. that the vessel desired to pass through the draw. At no time after the passage of the train over the drawspan did either of the bridge employees give the vessel any indication, other than keeping the drawspan closed, of further delay in the opening of the draw. If the draw was not ready to open at that time, the bridge should have informed the vessel of that fact either by radiophone or by an appropriate horn or whistle signal. The district court did not err in finding that the bridge was at fault and that such fault was a cause of the collision.

Finally, the Railroad claims that even though it was at fault, the vessel had the last clear chance of avoiding the collision. This claim is nothing more than an attempt to reargue the issue of the vessel’s negligence which has been resolved against the Railroad. Of course the vessel had a chance of turning aside in safety before reaching a point where collision was inevitable. However, the district court has found that her pilot, in permitting the vessel to approach that point, reasonably relied on the conduct of the bridge in doing so. The district court was acting within its province as the finder of the facts in refusing to apply the doctrine.

The decree of the district court will be affirmed.