This is an appeal from the unanimous decision of the Supreme Court of Canada, holding the appellants, the owners of the Norwegian Steamship Norefiord, solely to blame for a collision which occurred on 20th August 1942, between the Norefjord and the United States Steamship Alcoa Rambler (hereafter called the Rambler) the owners of which were respondents in the appeal. The case was tried in the first instance in the Exchequer Court of Canada, Admiralty Division, by the Hon. Carroll J., who held both vessels to blame, apportioning the blame as to three fourths to the Rambler and as to one-fourth on the Norefjord.
 The collision occurred in the Bedford Basin in the Harbour of Halifax, Nova Scotia. The Basin is a large expanse of water, a few miles in length and breadth, roughly rectangular or rather perhaps oval in shape, its greater length being from North to South. At the southeastern corner, there is an exit called the Narrows, leading into a channel which runs past the City of Halifax to the Atlantic. The water in the Basin is deep; in some places soundings are 100 feet or over. In 1942, it was an area much congested, as was natural in war time, crowded with vessels at anchor or in motion, estimated to number 100 or more, in addition to smaller craft, such as tugs, barges, launches and the like. There were recognised anchorages at various places, but under the stress of war conditions, vessels might find anchorage in almost any part of the Basin. The Fairway from the northern or north western area to the Narrows ran about S. 8. E. To the east of the Fairway there was a long stretch called the Degaussing Range running a little east of south, in which various operations connected with the degaussing of vessels were carried out under the Canadian Naval Authorities, Lieutenant Dyke, who gave evidence at the hearing, being in command.
 On the day of the collision, both the Norefjord and the Rambler were anchored at their respective berths, the Rambler at an anchorage about the north-vest corner of the Basin, and the Norefjord at an anchorage well down to the south on the west side. It happened that the two vessels quite independently started off on their different manoeuvres at about the sametime on the morning in question. The weather was fine and clear, there was no tide or current to speak of, and very slight if any wind. Each vessel was in charge of a duly qualified Halifax Harbour pilot. Each vessel was heavily loaded.
 The Rambler is a United States screw steamship of 5500 tons gross and 3381 tons net, register, 417 feet in length and 54 feet beam; she is fitted with turbine engines of 3000 horsepower, driving a single shaft with a righthanded screw. Her cargo partly consisted of United States Army warlike stores, but also included 1500 tons of bombs, a dangerous cargo making her an ammunition ship within Canadian War Regulations and the special practices of the port of Halifax. Her speed before she sighted the Norefjord was about 5 or 6 knots. The Norefjord belongs to the port of Oslo, and was at the time under requisition by the Norwegian Government, her cargo consisted of 4653 tons of sulphur. She had reciprocating engines. Her net tonnage is 1917 tons, her length 331 feet, her beam 46 feet. Her speed, on the morning in question was maintained at about 6 knots until just before the collision.
 Though both vessels started about the same time, their destinations were different. The Rambler intended to proceed down the Fairway to go through the Narrows and then by the channel into the Atlantic for gun practice. The Norefjord intended to make for the Range, entering it at the southern end. The Rambler's course was about S. S. E. magnetic. From her starting point, the Norefjord had to go some, what north of east to reach the Range. The Norefjord was pushed by a tug at her anchorage into an easterly heading in order to start. The Rambler started under her own steam. It was about 9.14 or a little before when the vessels sighted each other, at a distance it is estimated of about 1800 or 2000 feet. The evidence of the pilot of a Norefjord was that he intended to cut right across the Fairway in one manoeuvre to the Range, notwithstanding the intervening crowd of shipping. Obviously his course to the Range could not be straight because he had to progress not only in an easterly direction but also make some distance to the north, because the point at which be intended to come upon the Range was substantially north of the point from which he started. He intended to cut the corner, as it were, instead of proceeding east and then turning north. His approach, therefore, was what he was willing to describe as a "curved course." The evidence is that be was constantly swinging to port. It was a "curving course." That was not merely to avoid obstacles like the tug and barge which be met just before sighting the Rambler. Mere deviations from a straight course might be merely ordinary incidents of manoeuvre in passing through an area congested with shipping, but a deliberate departure from keeping a defined course would be a different matter.
 It may be noted here that neither pilot was working to compass courses or taking bearing or fixes. It was not necessary for the pilots to do so. They were manoeuvring in clear daylight, in full view of their destinations and of all the intervening shipping. But the situation was complicated by the number of vessels all over the Basin and the constant changes of direction necessitated in order to avoid vessels at anchor or moving. In these conditions it was difficult to get uninterrupted views ahead or around. All this greatly increased the danger of what the Norefjord was doing when she set out to out across the Basin at considerable speed right across the Fairway and in the track of outgoing and incoming chips and in the midst of other vessels moving about in so busy and congested an area. The conclusion of this appeal cannot depend on the nice assessment or comparison of points, angles, bearings or distances but on the broad features on which alone a decision can be reached. It has been accepted that the times in the rough scrap notes kept by the engineer of the Rambler are substantially accurate. They will be set out here in a moment. Before doing so, it may be convenient to mention a special complication which arises from the nature of the Rambler's part cargo of explosives. This put her in the category of an ammunition ship and brought her within the range of various provisions and regulations ruling in the port of Halifax. Ammunition ships were by law or at least by practice required to carry the red flag B of the International Code of signals and to be escorted in and out of the harbour by launches each carrying the flag B and also a second flag bearing the initials I. C., which signifies "Keep away." This was done in the case of the Rambler. On it was based a contention which was developed in the Supreme Court that there was a duty on the Norefjord which overrode the effect of Arts. 19 and 21 of the Regulations for Preventing Collisions at Sea and required the Norefjord to give way to the Rambler as contrasted with the duty attaching to the Rambler if Arts. 19 and 31 applied, to keep away from the Norefjord, which at all material times the Rambler bad on her own starboard side. Their Lordships as will appear later do not think that the Regulations in fact control the situation here. They will accordingly leave till a later stage of this judgment the consideration of the important questions fully discussed by the Justices of the Supreme Court on this aspect of the appeal. It is enough to say that the Norefjord must, in their Lordships' opinion, be taken to have been fully aware that the Rambler was an ammunition ship, by reason of the flags she and her escort were carrying, in addition to the specific warning or order given to her as she proceeded by the coxswain of the escort vessel.
 Their Lordships accept the times which appear in the Rambler's engine room rough record after the starting time at about 9 A. M. They are 9.14 slow ahead, 9.15 stop, 9.16 full astern, 9.17 full astern jingle, 9.17 full astern jingle. The collision occurred at 9.18. By jingle is meant an urgent ringing from the bridge to the engine room, repeating and emphasising the urgency of an order already given, ringing the engineer to put every ounce of steam into the execution of the order. It does not mean that the order is not being obeyed by the engineer. The master who gave the "jingle" on the verge of the collision saw the extreme danger then imminent.
 The order to stop at 9.15 was given because the Rambler had seen the masts of the Norefjord behind an intervening anchored ship and when they were seen to be indicating movement on a beading which might be likely to converge on the Rambler at an angle of about 2 points be decided to stop the engines. But it was not thought it was necessary to do more than stop the engines until the other vessel's hull came in sight round the stern of anchored vessel, or to take the very drastic measure of reversing which was a serious matter for a laden vessel like the Rambler weighing about 9,000 tons. Hence the interval of 1 minutes before the full astern order was given: it was given immediately the hull of the Norefjord was seen and appeared to be cutting across the Basin. The order was immediately carried out. But it is in evidence and accepted by their Lordships, confirmed as it is by the advice of the Nautical Assessors, that it takes 40 seconds in the case of such a vessel with turbine engines for a full astern order to take effect. This may explain the curious evidence given by one of the appellant's witnesses. Verge, the master of, the tug attending the Norefjord, that when the order was given the Rambler instead of going astern went ahead for 8 minutes. The trial Judge accepted that evidence and bated on it his finding that the order to stop was not acted upon at once. But in the Supreme Court this conclusion was convincingly analysed and rejected by Kellog, J., and their Lordships are in full agreement with that learned Judge that Verge's evidence, which was contrary not only to all probability but to the other evidence in the case, cannot be accepted. This seems to remove the sole ground on which the trial Judge attributed blame to the Rambler. Their Lordships are satisfied, that she reversed in good time. They will revert to that issue later. As things were then, even if the Norefjord bad the status of stand on ship under the Collision Regulations, a view which their Lordships as they will explain later do not accept, the time had come for the Norefjord to take action to avoid collision as required by her duty as stated by Sir Gorell Barnes in S.S. Albano v. Allan Line Steamship Company Ltd., (1907) A. 0. 193 at p. 207 : (76 LJ PC 33). Assuming the Norefjord had or thought she had the status of the stand-on ship the time had come when it was her duty to take action to avoid collision. But she did nothing to this end. Instead she continued to proceed at 6 knots as she had been doing up to that time and continued to do so until just before the actual impact. She had been compelled, she said, to starboard in order to avoid a tug and tow just before the Rambler sighted her, and then to port in order to pass under the stern of an anchored vessel. Thereafter, she said she put her helm to starboard to correct the port swing. It was then she said she gave the single blast, which helped to make it difficult for the Rambler to understand what the Norefjord was doing. It was this signal given in these circumstances which led the trial Judge to hold the Norefjord to blame to the extent of one fourth part. Indeed it must certainly have puzzled the Rambler to see the Norefjord swinging to port and at the same time hear her one short blast. In any event the Norefjord proceeded at a speed of 6 knots as she had been doing since she started from her anchorage. She went on doing so, notwithstanding the three short blasts sounded by the Rambler when she went astern, and notwithstanding the K or danger signal (one long, one short, one long), also sounded by the Rambler meaning "You should stop your vessel immediately." There was also the warning given from the escort vessel. Rand J. in the Supreme Court has calculated that the Norefjord went about 1,800 feet from the time the vessels came into full view until the collision. Their Lordships think this figure should be taken as substantially correct. There was abundant time for the Norefjord to. realise and discharge her duty to take action to avoid or at least minimise the effect of the Collision. She however did not change course or speed until just before the collision when it was too, late to avert the impact. The Rambler came in contact with the port side of the Norefjord about 215 feet abaft her bow. The blow was almost right angled, perhaps at 76 degrees. The Rambler was then almost stationary-moving at most at one knot but her great weight was sufficient to do the damage to the Norefjord, which was in the way of the engine room and stokehold. The abortive and belated efforts which the Norefjord made at the last moment could not affect fee result, though she dropped both anchors and went astern and then ahead and astern in order to help her swing to starboard. Lieutenant Dyke who had been waiting in his launch for the Norefjord has excellently plotted out the final movements. It appears that the Norefjord had changed 4 points from the bearing on which she started, which agrees with the swinging to port spoken of by the witnesses and was inevitable if the Norefjord was to achieve her carving course and reach the Range as it was intended. Lieutenant Dyke gave evidence that while he waited, she was constantly swinging to port.
 The appeal has been principally but not entirely argued before their Lordships on the navigational issue. In the Supreme Court the majority of the learned Judges who gave reserved judgments rested their decision on the status of an ammunition ship in the port of Halifax. Their Lordships, who have beard elaborate arguments on the questions of navigation and seamanship which are involved, think that the appeal can be sufficiently decided on these issues for reasons which can now be stated. They will later add dome observations on the important issues emerging because the Rambler was an ammunition ship.
10. In the forefront of the appellants' case it is stated that
"the main question in this appeal is whether the Alooa Rambler was under a duty within the meaning of Art. 19 of the Regulations for Preventing Collisions at Sea to beep oat of the way of the Norefjord which was at all material times on the starboard hand of the Alcoa Rambler and on a course crossing that of the Alooa Rambler from starboard to port."
Their Lordships, however, do not think that question correctly or sufficiently states the problem. If it be granted, as it must be that the Rambler had the Norefjord on her starboard side, there still remains the question whether the Norefjord was on a crossing course or on a "course,' at all. Their Lordships think this should be answered in the negative, but even if it were answered in the affirmative, there would still remain for consideration whether the risk of collision was not such as to put the duty of taking action on the Norefjord at the moment or practically at the moment when the ships first flighted each other, and the Norefjord did not discharge that duty. That is to say, if Arts. 19 and 21 ever applied, they were superseded by Art. 27. In either event the case will have to be decided on grounds of good seamanship and according to the common law of negligence. Articles 19 and 21 have always to be read together with Art. 27, which imposes an imperative obligation upon the stand on ship as soon as the necessary conditions are fulfilled. This principle is clearly stated by Sir Gorell Barnes P. in the Albano : (1907 AC 193) (supra) where (p. 207) he said of the master of the stand on vessel that "he must keep his course and speed up to some point (that is the time at which be must take some action), and then act, but the precise point must necessarily be difficult to determine, and some little latitude has to be allowed to the master in determining this." Article 27 is imperative :
"In Obeying and construing these rules, due regard shall be had to all dangers of navigation and Collision, and to any special circumstances which may render departure from the above rules necessary in order to avoid immediate danger."
11. Articles 19 and 21 presuppose as their essential conditions that the vessels must be crossing vessels and crossing so as to involve risk of collision. It is only when these conditions are present that the Articles apply and when they cease to be present the Articles cease to apply in the full sense that is disregarding for the moment Art. 27. But the problem does not depend merely on physical circumstances. As the purpose of Art. 19 and 21 is to impose a duty on the giveway ship to keep clear, that ship must be in a position to appreciate what the situation is and to know what the other ship is doing and whether it is on a course at all or if so on what course. In the present case, the Rambler was presented with two alternative possibilities in regard to the Norefjord; she might be making for the Narrows or some other place on the east side of the Basin or elsewhere; but the Rambler could not in the circumstances say which. If the Norefjord was going to the Narrows she was clearly not a crossing vessel, if she was making across the Basin that would be a different situation. When the vessels first sighted each other, the Rambler was left in doubt; even if, which is doubtful, she ever received reliable notice that the Norefjord was making for the Range, that notice was too late because in any view it did not reach her until she had stopped. The question whether the two vessels were crossing vessels within the article thus depended on the purpose of the Norefjord's movement, so far as that purpose could be inferred by the Master of the Rambler from what he saw. The issue always depends on the reasonable inference to be drawn by the ship which has the other on her starboard side as to the latter vessel's future course deduced from observation of her movement at the relevant moment, making due allowance for the nature of the locality where she is at that moment. So it is stated in the Albano : (1907 AC 193) (supra) at p. 205, in a quotation made by Sir Gorell Barnes from Lord St. Helier in The Pekin, (1897) AC 531 at p. 536 : (66 LJ PC 97). Thus there are two questions: (l) was the Norefjord on a crossing course, i.e., was she on a arise at all and if so was it a crossing course so as to be likely to involve risk of collision; (2) was the situation such that the Rambler, the give-way ship, could appreciate what the situation was so as to be able to act as the Rule prescribes.
12. It is difficult to define what is meant by "course" in this connection. It has been said that it does not mean "the actual compass direction of the heading of the vessel at the time the other is sighted"; so it was said by Lord Alver. stone C. J. in the Roanoke (1908) p. 231 at p. 239: (77 LJ P. 115) citing the Velocity, (1870) 3 PC 44 : (39 LJ Adm. 20) where it was held that the course of a ship going round a bend in a river could not be determined by her heading at the moment when she was at the turn of the bend. Lord Alverstone adds :
"A vessel bound to keep her course and speed may be obliged to reduce her speed in order to avoid some danger of navigation and the question must be in each case 'is the manoeuvre in which the vessel is engaged an ordinary and proper manoeuvre in the course of navigation which will require an alteration of course and speed : ought the other vessel to be aware of the manoeuvre which is being attempted to be carried out ?' "
It is only when the stand on vessel is aware or ought to be aware of that fact that the vessels are crossing vessels within Ss. 19 and 21 because that awareness is a necessary condition. This implicitly at least involves that prima facie any alteration of direction is inconsistent with the idea of a "course" within the meaning of the Regulations. These principles are illustrated by the Albano : (1907 AC 193 : 76 LJ PC 83) (supra) where the manoeuvre was being executed in order to pick up a pilot : other illustrations given were stopping and turning to get to a wharf, and the like. The test always is : was what was being done open and notorious to a seaman on the other ship in the ordinary course of navigation There is however a more general question which has to be considered in this case, namely whether what the Norefjord was doing was being on a "course" when she was making what the pilot called a curved course. The ordinary idea of a course, is a sufficiently constant direction of a ship on the same line or heading. This will enable a navigator when he sees the other vessel to know if she is on a crossing course. He can often only become aware of that if he can keep the other vessel under observation for sufficient time to ascertain if she is or is not changing her heading. In the open sea, this is the usual procedure as for instance.in the Otranto (1931) AC 194 : (100 L. J. p. 11). In that case the navigating officer of the Otranto, the stand-on ship, watched the give-way ship, the Kitano Maru; he first saw the latter vessel when 6 or 7 miles distant, a little before 8-30 : he took a compass bearing and then took another : at 8-40 finding she was not altering her bearing he sent for the Captain; then at 8-45 as her bearing still was unaltered, the situation was considered so dangerous as to justify and require the Otranto to take action for the other vessel. This was held to be correct to that extent, though she was the stand-on ship. But the helm action which ha took was the wrong action; she should have stopped and reversed. This case is cited here as an illustration of the critical test for inferring that the vessel is on a course that her bearing does not alter. It may often not be possible in narrow or congested waters; in the present case it could not be applied in regard to the Norefjord. If however it had been possible to watch her movement for some time, the manoeuvre would not have shown that she was keeping a steady course: her heading would have been altering to port. A curved or curving course, constantly changing under a port helm would not have enabled the give-way vessel to decide how to act for her. Accordingly of the two conditions that she should be on a course and that the give-way vessel should be able to ascertain that, neither was fulfilled. It may be that in crowded or congested areas, it may not always be possible to ascertain whether tie necessary conditions exist. In such cases Arts. 19 and 21 cannot apply. But wherever possible Arts. 19 and 21 ought to be applied and strictly enforced because they tend to secure safe navigation. It is for this reason that the principles laid down in cases like the Velocity : (1870 3 PC 44 : 39 LJ Adm. 20) (supra) the Echo (1917) P. 182 : (86 LJ P 121) and others have been found advantageous because they have enabled a wider scope to be given to the crossing rule, even in cases where in a strict sense there is deviation from the ship's course. They also illustrate the importance attached to the necessity of attributing to the give-way ship actual or imputed knowledge of the situation.
 In the present case there was no obvious or ordinary manoeuvre which would give knowledge to the Rambler so that quite apart from difficulty inherent in a curved course, the case could not be held to be a case of crossing vessels because the necessary knowledge of the situation could not be ascribed to the Rambler.
 In any event, the Rambler took correct action at the proper moment. There was
no real question here of helm action being the best course. As to stopping or reversing the only question raised was whether she should have reversed when she merely stopped. Their Lordships have indicated their conclusion that stopping and not reversing was the correct step at that moment. On that issue, which is a matter of good seamanship, their Lordships have taken the advice of the Nautical Assessors, whose opinion coincides with their own. They think the Rambler free from blame. She took appropriate action at the appropriate moment as a matter of good seamanship. Their answer to their Lordships' question is
''In our opinion Rambler took a reasonable and prudent precaution in stopping her engines on sighting the masts of a ship which was under way and in our opinion good seamanship did not require her to reverse her engines at this stage."
This disposes of the suggestion that the Rambler should have reversed at the earlier stage, which is the only serious objection suggested against what she did. As to the Norefjord the advice of the Nautical Assessors is that when the Norefjord in the circumstances saw the Rambler "good and careful navigation called for the Norefjord to stop her engines and thence proceed with caution until she saw her passage clear. This she could have done with perfect safety in the weather prevailing,"
instead she was guilty, in the circumstances stated above, of gross and palpable negligence, whether tested by the Regulations or according to the ordinary rules of negligence. She seems to have gone on in complete oblivion of the presence of the Rambler notwithstanding all the warnings and signals that were given to her. 'Their Lordships think the Norefjord alone to blame and liable for the whole damage.
 Their Lordships accordingly agree with the unanimous conclusion of the Supreme Court that the Norefjord was solely to blame.
 But they do so solely on navigational grounds and have not thought it necessary to examine, or come to a decision upon the important questions of the law and practice of the Fort of Halifax on which the majority of the Supreme Court Judges largely based their decision.
 During the First World War, the Bed-ford Basin was crowded with shipping (as it has been during the last World War), including many ammunition ships. In 1917 one of these latter was run into while in the Basin and as a result of the collision blew up. The explosion wreaked the City of Halifax and killed about 17,000 of the inhabitants. Early in the last World War, a Canadian Order in Council DC 2412, issued under the Canadian War Measures Act, provided for the Governmental regulation of navigation within Canadian waters, and among other things by cl. 1 required every vessel in those waters to comply with such orders as to the navigation and mooring of vessels as may be issued by the Minister and to obey any orders given either by signal or not by any officer in command of any of His Majesty's Canadian ships. By S. 38 (a) of certain regulations issued pursuant to the above Order in Council, ships entering harbour and carrying explosives were to be escorted by Naval Craft flying letters I. C. and later when proceeding to a berth in the Basin. "All ships seeing the Naval vessel flying International are to keep well clear." There was no express provision for ammunition ships proceeding out of the Basin as the Rambler was, but there was a practice, well known to the pilots concerned, including the pilot of the Norefjord, imposing the same requirements on vessels proceeding outwards as on vessels proceeding inwards.
 In these circumstances, the Chief Justice in the Supreme Court held that these local regulations, coupled with the known and terrible risks of an explosion, constituted " special circumstances " within Article 27 of the Regulations for Preventing Collisions at Sea "which may render a departure from the above rules necessary in order to avoid immediate danger." He held the Norefjord solely to blame for not keeping out of the way of the Rambler. on account of her dangerous cargo. He held that the Rambler was entitled to stand on her status as an ammunition ship and assume that the Norefjord would keep out of her way. Kerwin J's. concurring judgment gave no reasons Tascherea, J. did not invoke Regulation 33 (a) but held the Norefjord solely to blame for faults in navigation, specially emphasising her wrong single blast signal. Rand, J. based his decision in respect of the liability of the Norefjord on the local regulations, as being special rules of ft local authority within Art. 30, Collision Regulations and hence paramount to Arts. 19 and 21. His review of the navigation was however definitely in favour of the Rambler. Kellog, J., whose criticisms of the findings of the trial Judge have already bean referred to, was of opinion that the Norefjord was not on a course when she first saw the Rambler. He held that the Naval Escort vessel gave an "order" to the Norefjord which constituted under S. 33 (a), of the regulations made under the Order in Council an absolute obligation to keep clear of the Rambler; in respect of her seamanship he found nothing to blame.
 Their Lordships do not think it necessary either to affirm or disaffirm the opinions expressed by the Supreme Court on the local regulations or the law and the practice in the Port of Halifax. They may however observe that even apart from these regulations and practice the balance of the opinions delivered by the Court is generally, so far as concerns navigation, in favour of their Lordships' view as to the culpability of the Norefjord.
 They will humbly advise His Majesty that the appeal should be dismissed and the judgment of the Supreme Court of Canada affirmed. The appellants will pay the costs of the appeal.