Neuchatel Swiss General Insurance Co Ltd & Ors v Vlasons Shipping Inc [2001] VSCA 25

2. This appeal arises out of a judgment of a judge of the Trial Division which held that the appellant insurers were liable to the respondent ship owner, pursuant to a hull (time) companies combined marine insurance policy. The policy insured three vessels for the period 16 March 1991 to 16 March 1992 in far-east waters, and the judgment held the appellants liable for the loss of one of the ships, the Biyayang Ginto, insured at a value of $US1,200,000, which went down with all hands in a typhoon in the South China Sea on 16 August 1991. The issue raised by the proceeding is whether the insurers were entitled to rely upon what was called a "typhoon clause", which was inserted by way of endorsement on the policy in the following terms:

"NOTWITHSTANDING any provision of this policy to the contrary, it is hereby expressly stipulated and made a condition herein that the insured vessel shall not be allowed to sail or put out of port when there is a typhoon or storm warning either at the port of destination or between the said port. Violation of this condition shall render the policy NULL AND VOID. HOWEVER, should the vessel have sailed out of port prior to such warning, this warranty shall no longer apply."

The question which arose in the proceeding was whether the appellant insurers were entitled to deny liability under the policy because, as they alleged, the vessel had been "allowed to sail or put out of port when there [was] a typhoon or storm warning either at the port of destination or between the said port" [sic]. The clause was inelegantly drafted, as the appellants conceded, and it would seem either that the particular clause had never been used previously or, that even if it had, none of these parties, nor the very experienced counsel who appeared on their behalf, had come across such a clause before or since. It is not entirely clear whether "port of destination" should read "port of sailing" or "port of departure", but it was otherwise agreed between the parties that the otherwise unintelligible expression "between the said port" should be treated as meaning "between the port of sailing and the port of destination".

7. The learned judge held that the condition did not apply, so that the respondent owner of the ship recovered on the policy in an amount of $US1.2 million together with interest, which the judge calculated in accordance with published US prime borrowing rates, to which he added a further one per cent. There were, in substance, three issues raised on the appeal. The first concerned the kind of storm and typhoon warnings which would activate the condition; the second was whether the master of the ship had to be shown to be aware of the existence of the kind of storm or typhoon warning referred to in the condition; and the third related to the calculation of interest and in particular to the right of the judge to add a further one per cent to the published rates.

10. The first issue raised directly the construction of the condition inasmuch as the appellants had sought to invoke it in relation to warnings issued by the Royal Observatory of Hong Kong ("ROHK") and the Japan Meteorological Agency (in Tokyo) ("JMA") which the evidence clearly showed had been issued relating to a tropical storm which later became a typhoon. These had been issued by those agencies not long before the subject vessel, the Biyayang Ginto, had put to sea on 14 August 1991 from the Port of Beihai on the south coast of China, on its way to its intended port of destination at Hong Kong over two days' sailing to the east of Beihai. At the time of these warnings the storm, later to become a typhoon named "Fred", was several hundred miles south of the vessel's proposed line of voyage but both agencies warned that Fred would cross the proposed line of voyage within the succeeding 48 hours. For various reasons, to which I shall return, the master was unaware of these warnings when the ship left port. So it happened that, as the ship turned east to travel through the strait which separates Hainan Island and the Chinese mainland, what by then had become a typhoon struck the vessel, causing it to sink with all hands lost, although the point at which the vessel sank was a considerable distance to the west of the paths which both agencies had expected the storm to travel.

13. The issue therefore was whether storm warnings of this kind were intended to relieve the insurance companies of their obligations under the policy. The learned judge held that they did not, for he held that warnings of the kind described in the condition had then to affect either the port of destination (Hong Kong) or the vessel's proposed line of voyage from Beihai to Hong Kong, in each case at the time that the warning was given. In other words, the condition referred only to an existing danger, of which relevant agencies were giving warnings, at the time immediately before the vessel left port. The appellants have contended, however, that such an interpretation is both unrealistic and too narrow in that a warning must refer to a storm or typhoon of a kind which the warning states is likely in the future to affect a ship's voyage to its destination. As in the present case both warnings were to the effect that the typhoon was forecast to cross the line of the vessel's journey at some stage during its voyage to Hong Kong, so those warnings were of a kind sufficient to require the vessel to desist from sailing if the condition avoiding liability and the policy was not to affect the insured.

16. Secondly, the learned judge found that, by reason of the expression "shall not be allowed to sail ...", the typhoon clause did not operate to deny liability unless the master had knowledge of a relevant warning. In the circumstances his Honour was unable to find that the existence of the storm or typhoon warning was brought to the attention of the master of this vessel or anyone in authority on the ship before it left port. The appellants challenged this finding also in the first place on the ground that, in a condition such as the present, knowledge by the master or anyone else in authority (whether on the ship or on land) was not relevant and the burden was on the master or ship owner to ascertain the weather conditions and any relevant storm or typhoon warning. Next, it was asserted that, if there were a requirement as to knowledge, it would have been sufficient that the warning came or should have come to the attention of the master or anyone who had authority to control the ship's movements before it left port, so that proof of actual communication to the master was not relevant. Finally, it was argued that as a matter of fact the proper inference was that the master received a warning transmitted some time earlier to the local agents in Beihai, which warning had come to his knowledge before the vessel left port.

19. The final issue on this appeal arose out of a matter which was argued separately and led to a judgment by the learned judge given on 20 November 1998. That issue raised the extent of the discretion of a trial judge in awarding interest on a sum awarded at trial. Two sets of rates were tendered to his Honour for the purpose of this argument, the judge preferring those tendered by the respondent over those provided by the appellants from the Reserve Bank of Australia. However, the issue now is not as to which of those sets of rates was correct, for the appellants accept that the judge was entitled to accept those provided by the respondent. The issue is only as to whether the judge was entitled to add one per cent to the rates so proved in evidence, and, in substance, whether the adding of one per cent exceeded the power of a judge at trial. The appellants contend that, where evidence is given and accepted as to relevant rates, as in this case on a debt payable in US dollars, the learned judge was not entitled as a matter of discretion to add a further one per cent to the rate so proved in evidence.

22. It is necessary to set out some of the circumstances surrounding this dispute. The respondent company, Vlasons Shipping Inc. ("Vlasons") is a Filipino company and was at all material times the owner of the vessel. It was, however, managed by a Hong Kong company, Northern Star Shipping Co. Ltd. ("Northern Star"), of which Mr Danny Shan was the director responsible for the management of the vessel in 1991. In turn, Northern Star engaged a number of agents at various ports to which its ships voyaged and in Beihai (also known as Pei-Hi) the relevant agent was a company known as Penavico. Its representative so far as the Biyayang Ginto was concerned was the deputy general manager of a subsidiary of Penavico whose name was Ruan Ruijie, who gave evidence at the trial.

25. The vessel was one of three ships insured by the respondent under a Companies Combined Policy of Marine Hull and All Risks Insurance, No. 3065, taken out with the appellant underwriters. The period of insurance was from 16 March 1991 to 16 March 1992, the particular vessel being insured under the policy for $US1.2 million. The policy contained the typhoon clause set out in paragraph [1] above.

28. The Biyayang Ginto was a small vessel (2997 tonnes and 101 metres in length) engaged in carrying goods in the South China Sea and on the present occasion it put into Beihai to load a cargo of Kaolin clay. That port is in the Gulf of Tonkin (Beibu Wan Gulf) and is a short distance west of the Luichow Peninsula on the south coast of China, immediately north of Hainan Island. The vessel arrived at the pilot station at Beihai Harbour at about 2200 hours (Hong Kong time, which will be used for convenience throughout this judgment) on 10 August 1991 and on the following day, in accordance with local practice, a joint entry inspection was conducted at about 1230 hours. Mr Ruan participated in this inspection together with the harbour master and other government officials. Eventually the vessel berthed alongside a wharf in the harbour at about 1800 hours on 12 August 1991. Loading of the cargo of clay commenced then and was completed at 0800 hours on 14 August 1991. Then at 1030 hours that day a joint exit inspection of the vessel was conducted and concluded not long afterwards, again with Mr Ruan and representatives of the same relevant authorities. It did not leave port immediately, apparently because it was experiencing some engine problems. It seems, however, that in Chinese ports at the time, once a ship had been cleared, no access from the shore was allowed.

31. During the time the vessel had put into the Port of Beihai and was loading its cargo a storm had been building up in the South China Sea not far from the northern tip of Luzon Island but not much further than 400 nautical miles from the southern coast of China. Even by 12 August the storm had become of sufficient severity to result in tropical cyclone warnings being issued by ROHK and not long afterwards similar warnings were coming from JMA's office in Tokyo. A number of definitions are given by the World Meteorological Organisation, two of which were accepted as being relevant for the purpose of the present clause. They were:

"STORM WARNING Meteorological message intended to warn those concerned of the occurrence or expected occurrence of average wind speeds in the range of 48 to 63 knots or windforce 10 to 11 in the Beaufort scale. TYPHOON WARNING Meteorological message intended to warn those concerned of the occurrence or expected occurrence of the mean wind speed of 64 knots or higher, or wind force 12 in the Beaufort scale."

The significance of these warnings is that they formed the basis for various radio transmissions from meteorological offices around the world intended to warn shipping, in particular, of the relevant risks. Vessels of the size and class here in question ordinarily had radio rooms with equipment capable of receiving messages of this kind and it was ordinarily the duty of radio officers on these vessels to monitor messages which were, of course, frequently transmitted as part of ordinary weather forecasts or in conjunction with them.

36. For example the ROHK issued daily weather reports for the region including the waters between Beihai and Hong Kong and there were more frequent reports when storms were in the relevant areas. The information contained in the ROHK reports was used for preparation of marine forecasts and tropical cyclone warnings. For this purpose a tropical cyclone is a general term used to describe a tropical low pressure system with organised convection and definite cyclonic wind circulation. The marine weather forecasts received from the ROHK were broadcast to shipping in morse code from the Hong Kong coast station at three specified times each day, namely at 0918, 1518 and 2118 hours. When a tropical cyclone warning was passed on to the Hong Kong coast station it was transmitted to shipping in morse code immediately on receipt from the ROHK and thereafter at 18 minutes past the hour each hour for three hours. More often than not it was then updated and the same pattern continued thereafter.

39. Weather summaries and cyclone warnings were also issued at the time by the JMA, which were likewise broadcast to shipping by morse code at regular intervals by the meteorological broadcast station known as JMC. General weather forecasts and warnings were issued every six hours by the JMA and transmitted at 0200, 0800, 1400 and 2000 hours. If there were warnings, then it seems that they were repeated additionally 2 hours 48 minutes after the primary report and warning and again 3 hours and 48 minutes after the earlier report.

42. In each case the weather bureaux attempted to forecast the expected path of the storm or typhoon 24 and 48 hours later. So, as might be expected, cyclone warnings did not confine themselves merely to giving information as to the current position of a storm or typhoon but attempted to warn shipping of the expected course of the storm or typhoon.

45. Thus, in the present case, when the warnings first began to be given the vessel was more than 400 miles away from the eye of the storm while it was loading its cargo in the Port of Beihai. Even at that stage, however, it seemed reasonably clear that unless something unusual occurred the storm would come closer and closer to the proposed line of voyage of the vessel to Hong Kong. The various reports contained information as to the direction, essentially west north-west, and the speed, between four and six knots, of the storm at that time. In the following few days it moved much closer to the coast, although there was a relatively small difference (about 39 miles) as to the precise eye of the storm as reported by the two primary agencies, the ROHK and the JMA.

48. One tropical cyclone warning, seen by the parties to be relevant, was issued by ROHK at 0500 Hong Kong time on 14 August 1991, which read:

"AT 132100 GMT, TROPICAL STORM FRED (9111) WITH CENTRAL PRESSURE 990 HECTOPASCALS WAS CENTRED WITHIN 60 NAUTICAL MILES OF ONE EIGHT POINT NINE DEGREES NORTH (18.9N) ONE ONE EIGHT POINT THREE DEGREES EAST (118.3E) AND IS FORECAST TO INTENSIFY GRADUALLY AND MOVE WEST-NORTHWEST AT ABOUT 6 KNOTS FOR THE NEXT 24 HOURS. MAXIMUM WINDS NEAR THE CENTRE ARE ESTIMATED TO BE 40 KNOTS. RADIUS OF OVER 33 KNOT WINDS 60 NAUTICAL MILES. RADIUS OF OVER TWO METRE WAVES 200 NAUTICAL MILES. FORECAST POSITION AT 142100 GMT ONE NINE POINT EIGHT DEGREES NORTH (19.8N) ONE ONE FIVE POINT NINE DEGREES EAST (115.9E). FORECAST POSITION AT 152100 GMT TWO ZERO POINT SEVEN DEGREES NORTH (20.7N) ONE ONE THREE POINT SIX DEGREES EAST (113.6E)."

That indicated that the tropical storm was by then on a similar latitude to Hainan Island but still some hundreds of miles to the east. On the other hand, the forecast position of the storm at 0500 hours two days later, when the vessel was expected still to be at sea, was slightly north of the latitude of Hainan Strait and approximately two-thirds of the distance from that strait to Hong Kong, only a short distance from the expected course of the vessel to Hong Kong.

53. A not dissimilar warning apparently had been given by the JMA at the same time describing the storm as "tropical storm 9111 Fred" with maximum winds of 45 knots near the centre and with a slightly different forecast for the times 24 and 48 hours ahead, suggesting a track only slightly north of a due westerly direction. It appears that the warnings given at this time by both bureaux were seen to be important in that a copy of a weather chart issued by ROHK was sent by fax to Northern Star Shipping at Hong Kong on 14 August where Mr Shan saw it. He did not take any immediate steps but at 12.22 that day he sent a telex to Penavico in Beihai asking as to progress and also requesting Penavico to pass to the master the following message:

"Dear Captain, Understand there are [sic] typhoon Fred approaching which kindly take extra cautions and always consider safety first. Regards - Norstarco."

There was much dispute at the trial as to whether this message was taken to the wharf and handed to somebody on the vessel, which Mr Ruan claimed he had done but which the trial judge specifically held he could not accept. Whatever Mr Ruan claimed to have done, he certainly conceded that he was aware of the typhoon warnings while the vessel was still in port, having received them from both the harbour authorities and the local radio and television stations.

58. The claim as to the telex was important inasmuch as it seems that in practical terms the only way the master could have received any storm or typhoon warnings while in port in Beihai was by the local agent's sending or bringing them to him. This was because, at least as was accepted and as the trial judge found, when ships were in Chinese ports, their radio rooms and the equipment in them were sealed up by government authorities, seemingly to prevent transmissions being made from the port. Although ships such as the Biyayang Ginto also had a VHF receiver on the bridge, that could only receive local transmissions and it was not suggested that there were any local warnings transmitted on VHF, at least of a kind which the Filipino but English-speaking master of the vessel could understand. As I have said, the primary sources in the area for weather reports and warnings were ROHK and JMA, each of which could only be received on the ship's main radio. Although the regulations did not prohibit the receipt of messages by radio, the practical effect of sealing the radio and of the local practice of jamming foreign radio signals, as the judge found, was to deny the master and the crew of the vessel direct access to the relevant weather forecasts during the time the vessel was in port in Beihai.

61. Of course, at this time Northern Star had access to all the relevant forecasts, especially those issued by ROHK. It had received by fax a copy of the 0500 hours weather chart containing references to cyclone Fred which Mr Shan read shortly before 9 o'clock that morning. It was probably this that provoked him to send the telex message which he wished to be relayed to the ship's master. It will be necessary to return to the fate of that message and the evidence of Mr Ruan.

64. Although the vessel had been cleared for sailing by late-morning on 14 August 1991, the engine trouble had delayed it so that it did not leave port until some time between 1730 and 1800 hours that day. However, during the day the storm had been placed in a higher category. At 0800, JMA had upgraded the storm to a severe tropical storm with an estimated wind intensity of 50 knots near its centre. By 1400 ROHK also described the storm as a "severe tropical storm" with winds near the centre of a similar intensity. The final tropical cyclone warning issued by ROHK before the vessel sailed was issued at 1700 hours. It took an almost identical form to the warning previously set out[1] except that it now described the storm as "severe tropical storm Fred". The central pressure was 980 hectopascals with maximum winds near the centre estimated at 55 knots. The radius of 47 knot winds was 60 nautical miles. The storm was then centred at 19.8 degrees north and 117.0 degrees east, somewhat closer to Hong Kong and the coast although still a few hundred miles away. It was now moving at an estimated eight knots and the two forecast positions were now: at 1700 on 15 August it was estimated to be at 21.1 degrees north and 113.8 degrees east; and for 1700 on 16 August it was estimated it would be at 22.7 degrees north and 110.8 degrees east. The first position would therefore have been not far from the projected line of voyage of the vessel and the second position would have the storm over the coastline of southern China to the west of Hong Kong. The equivalent JMA warning, described as a "typhoon warning", although the storm was still characterised as "severe tropical storm 9111 Fred", was issued at 1400 hours. Again it was estimated to be 55 knots near its centre but moving west north-west at seven knots. Its estimated position was not greatly different from that estimated by ROHK but its forecast position for 1400 hours on 15 and 16 August was, as to 15 August, 20.5 degrees north and 114.0 degrees east "with an uncertainty of 90 miles radius", and at 1400 hours on 16 August, at 21.0 degrees north and 109.5 degrees east with an uncertainty of 180 miles radius. The two positions would have taken the storm somewhat further south than the path forecast by the ROHK warning, the first position being about 50 miles to the east of the estimated line of voyage of the vessel, but the second position would have had the typhoon crossing the Chinese coastline on the Luichow Peninsula immediately north of the Hainan Strait and thus a little north of the vessel's line of voyage.

67. As the judge found the facts, however, the master knew of none of these storm warnings. So it was that the vessel left Beihai late in the afternoon of 14 August 1991 bound for Hong Kong via the Hainan Strait. It is understood that after passing through the Hainan Strait the master intended to take the vessel north-east to Hong Kong, only about 300 miles away from the strait, although it was expected that the voyage in all would take the vessel approximately two days. From the start of the voyage, however, it was necessary to steam some 75 miles almost due south and then wait for clearance to enter the strait rather than to circumnavigate Hainan Island. It seems that the vessel arrived at the anchorage at the western end of the strait at about 1130 hours on 15 August, the master advising Penavico at 1359 of this fact. He still expected to pass through the strait at 0500 hours on 16 August and therefore to arrive at Hong Kong at 1500 hours on 17 August 1991. It was necessary to obtain clearance from the relevant authorities before entering the strait, although it was not necessary to be taken through by pilot.

70. It was while the vessel was waiting in the anchorage that the storm, now categorised as a typhoon, struck. Unfortunately, at this stage it had grown in intensity and had taken a track somewhat to the south of both of those forecast by ROHK and JMA, though still first hitting land on the Luichow Peninsula not many miles to the north of the strait and then following the course of the strait itself. The master sent out an SOS at 0800 hours on 16 August, when the eye of the storm was less than 40 miles away, but the vessel was never heard of again. The bodies of five sailors were recovered and identified and so it was known that the ship had sunk apparently with the loss of all 24 hands, still not far from the entrance to Hainan Strait. The typhoon itself eventually crossed the Vietnamese coastline, in all causing over 13 other deaths, destroying hundreds of homes and causing many millions of dollars damage.

73. If, as the learned trial judge held, the operation of the condition required that there be a typhoon or storm warning in existence either at the port of destination or between the port of sailing and the port of destination, then the evidence clearly showed that there was no warning of a storm at or about 1730 or 1800 hours on 14 August 1991 affecting either Hong Kong or any point on the route of the voyage contemplated by the master of the vessel from Beihai through the Hainan Strait to Hong Kong. On the other hand, if the appellants' contention as to the proper interpretation of the condition be correct, then, on the basis of the severe tropical storm moving in the manner forecast by the ROHK warnings, the storm would have crossed the proposed line of voyage of the vessel at about 0200 hours on 16 August 1991 and on the basis of the forecasts in the JMA warnings the storm would have crossed the same line at about 2300 hours on 15 August. Although there were a number of alterations in the proposed voyage times, it seems that for relevant purposes the estimated time of arrival in Hong Kong of the vessel was 1500 hours on 17 August 1991, a time substantially after the likely crossing of its voyage path by the storm on the basis of the forecasts of either ROHK or JMA.

76. As to the policy itself it is sufficient to say that it covered the vessel and two other vessels within "far east waters" and that it incorporated the Institute Time Clauses - Hulls of 1/10/83, as varied by the policy. The particular condition was incorporated by a provision in the schedule which referred to the "Conditions of Insurance" as being "as per schedule attached". That attachment was headed "INSURING CONDITIONS ATTACHING TO AND FORMING PART OF POLICY NO. 3065", in which it was simply stated, among a number of other provisions: "Typhoon clause (as attached)". The typed clause, in the form appearing in para.[1], was then separately attached.

80. As described above, the learned trial judge effectively held in favour of the respondent on two grounds: first, upon the basis that the warnings referred to in the condition had to be warnings of an existing typhoon or storm at the relevant places, there was neither typhoon nor storm (or any warning thereof) at any point close to Hong Kong as the port of destination or close to the line of voyage proposed to be taken by the vessel from Beihai to Hong Kong; and secondly, upon the basis that the condition required that the vessel be allowed to sail by the master with knowledge of

a relevant warning, the master of the Biyayang Ginto had not been shown to have had such knowledge.[2]

84. In my opinion, the judgment of the learned primary judge can be justified on the second ground he put forward. The use of the expression "shall not be allowed to sail or put out of port" connotes the need for the person in command of, or, arguably, otherwise in control of, the vessel to make a conscious decision to sail or put out of port and it connotes also in the context a decision in the face of knowledge that there is a typhoon or storm warning affecting the relevant line of voyage or destination port. That would be sufficient to deal with the issue of liability on this appeal, but I think in the circumstances it is appropriate also to look at the first matter argued on the part of the appellants, namely whether warnings of the future course of a storm came within the clause. To that issue I shall return in due course.

87. The construction to be given to the condition should be one appropriate to a commercial document of the formal kind represented by a marine insurance policy so that, although it should be given a sensible business meaning[3], it has also to be viewed in the light of the fact that it forms a condition to a policy having a specific effect on the parties' rights, in particular the insured's rights as affected by the second sentence of the first paragraph of the condition which made clear the fact that violation rendered the policy void. Nevertheless, a proper attempt must be made to give effect to the condition. Already the parties have accepted that the expression "between the said port" cannot be viewed as merely meaningless and thus struck out. Some words must here be read in, as they have conceded, and on this occasion it would seem that the expression should be read as "between the port of departure (or sailing) and the port of destination". If the learned judge were correct in his view that the relevant warnings must exist at the time of sailing then there might also have been some justification for correcting the expression "the port of destination" to "the port of departure (or sailing)". An existing warning would surely be more appropriate for a storm affecting the beginning of the voyage rather than its end, which might be several thousand miles away. Secondly, the subsequent expression "between the said port" (in the singular, it should be noted) might more naturally be amended by simply adding the words "and the port of destination". Of course, even the varied expression could not mean literally the direct line between the two ports; rather, as again appears largely to have been accepted, it must refer to the proposed line of voyage in so far as that can be identified before the time of sailing.

90. The respondent put forward a large number of further difficulties in giving the condition a precise meaning, each based on the premise that a condition with this effect ought to be read strictly and that any uncertainty ought to be resolved in the insured's favour. One may accept that such an approach may be justified in cases of genuine equivocation or uncertainty[4], but nevertheless a deliberately inserted condition of this kind ought to be given some meaning.

94. I turn first[5] to the question whether the appellants are correct in asserting that, on the proper construction of the policy, knowledge of the master or any other person allowing the vessel to sail need not be established for the condition to operate. The learned trial judge held that the expression in the condition "shall not be allowed to sail or put out of port" connoted that a decision must be made as to allowing the vessel to sail and that the only purpose of the words was to require that the person making the decision had the relevant knowledge of a typhoon or storm warning as described in the condition. The expression "be allowed to" certainly seems an unnecessary addition if the only question is whether there was in existence at the relevant time of sailing a proscribed typhoon or storm warning then affecting the expected line of voyage or destination port. If it were not for other factors, it might be thought that the expression was mere surplusage not intended to affect the interpretation of the condition. That interpretation certainly would have the consequence, as the appellants argued, that, if the master or other person responsible negligently fails to make enquiries about typhoon or storm warnings, then the insurers are not protected from the consequences of that negligence although it may lead to the loss of the ship. Whether negligence is relevant or not in the construction of any condition must depend upon its proper construction but, although it may be putting a premium on carelessness, it could not be said to lead to any deliberate flouting of the condition's requirements for it would not permit any deliberate decision to go to sea while such a warning was in existence and likely to affect the voyage.

97. The appellants insisted that the interpretation favoured by the learned judge would make it hard for the insurers to establish a deliberate decision to go to sea in the face of knowledge of a typhoon or storm warning. They said that the master would not be there afterwards to answer questions for he would have gone down with the vessel, as would all the records including those relating to warnings. That seems to me to be an overstatement for, although the master and all hands were lost on this occasion as they must be on others, it does not follow that in the case of every sinking the master is not rescued or does not otherwise survive. Moreover, those who provided the information to the master, if that be relevant, would frequently have a record of that. It was said that although there was a record of certain telexes being sent to the master while he was at sea, there could not be any proof that he received those messages. Again, I think that overstates the case for there were notations which indicated that the messages were received or not received, as the case may be, and in any event, absent unusual factors, the sending of a message in those circumstances might fairly lead to an inference that it was received. More importantly, however, the relevant warning must have been received by the master before he left port and there would ordinarily be better records of any relevant matters at that time. For example, in the present case the appellants sought to prove that a telex was hand delivered containing some element of warning and that might have been established, if it were not for the fact that the witness was held to be unreliable. Although there may be difficulties because the insurers bear the onus of proof, it ought not to prevent the clause being construed as requiring knowledge, if that is the appropriate way to interpret it. Moreover, the appellants' argument failed to take account of the fact that the vessel, as might others affected by a similar condition, remained in radio contact with its agents right up to the time of the SOS message. It cannot be of consequence for this purpose that the particular messages did not prove that the master had knowledge, for they might otherwise have provided evidence if knowledge existed. In the present case, however, as will be seen, there was no such evidence and the inference to be drawn, rather, is that the master did not have any such knowledge.

100. Whatever be the difficulties in the way of the insurers, the assumption behind the clause may well have been that the master or other persons responsible would naturally enquire about weather conditions before leaving port. It is hard to believe that a competent master would not have made enquiries as a matter of course especially at that time of year in the South China Sea. One is rather inclined to believe that he did make some such enquiries but was limited because of the restrictions placed on the use of the ship's radio, and thus he became dependent on such information as was passed to him from the shore or which he could get from local Chinese VHF radio stations. Mr Ruan did indeed concede that he had made some enquiries but the learned trial judge found as a matter of fact that he had not passed them on to the master. It should not be forgotten that at that time of day, that is, immediately before the final inspection was completed, ROHK had not yet upgraded the storm to a severe storm and JMA had done so only two hours earlier. Is it possible that the local weather station gave no intimation of the storm which at that time was closer to the Philippines than to China? It is of little consequence, for the judge found that the master did not have knowledge of the message.

103. There is a further reason why I consider that knowledge is a requirement of this condition, over and above giving a fair and proper meaning to the words "be allowed to" already discussed. That reason is that, unless there were some requirement for knowledge, the condition would become impractical having regard to the number of meteorological agencies which might issue forecasts and storm warnings in languages both familiar and unfamiliar to the master of the vessel. The requirement of knowledge overcomes the many difficulties otherwise put forward in relation to the proper construction of the condition by the respondent.

106. There are certainly apparent difficulties in giving a precise interpretation to the clause. They must fairly be acknowledged but a commonsense approach will overcome most of them. Some of the difficulties raised by the respondent in fact are equally applicable to the interpretation given by the trial judge and advanced by the respondent itself, or, if not equally applicable, pose similar problems. Nevertheless, if all those difficulties were added together the condition would be meaningless and that is an answer which the respondent never advanced on this appeal.

109. One particular difficulty which applies to the condition in the present context, but could equally apply in many other contexts, is that raised by the question as to which meteorological authorities' storm warnings are referred to. On one view, one favoured by the appellants' arguments, it is of no consequence from whom the warning comes so long as the danger will arise at the relevant time. They sought to overcome any apparent difficulty by asserting that in practical terms on the present occasion only the ROHK and JMA warnings need be looked at, but it is not so easy to overcome. Doubtless, the Hong Kong and Japanese bureaux are recognised and highly respected in the region but in somewhat different circumstances the same answer could not be so easily given. If the ship had been Chinese or its master Chinese-speaking, would it only have been necessary to have regard to forecasts and warnings given from Chinese meteorological sources? At one stage in the case such warnings were relied on and Mr Ruan conceded that he was aware of local warnings given in the Chinese language. If the ship had been close to the Philippines, would it have been necessary to have regard to warnings issued by the Filipino national bureau? What would happen if the warnings were not identical in each of the languages capable of being understood by the master? What if the master did not speak English of a sufficient standard to understand storm warnings but could understand only those given in his own language and perhaps in one other language? Likewise, in different parts of the world, and even in "far east waters", however that be defined, there would be a large number of agencies operating in a relatively small area each of whom might put out warnings and forecasts of different kinds. Would the condition operate if only one of those many warnings affected the line of voyage of the vessel, even if all others did not?

112. In my opinion, this difficulty can only be overcome if one accepts the argument put on behalf of the respondent, namely, that for the purpose of this condition it will only operate if the master or other person responsible for allowing the vessel to sail was aware of the relevant warning. The condition here could have applied if the master, with the relevant knowledge, "allowed" his vessel to leave port. That seems the only sensible way of resolving the many difficulties which can arise from the variety of meteorological bureaux who do issue warnings to shipping both in the South China Sea and elsewhere. Otherwise, the condition would operate where a bureau put out a seemingly relevant warning of which the master was unaware because either the master considered the bureau to be less reliable or he could not understand its forecasts and warnings. On the basis that knowledge is a requirement of the condition, then the objection as to the numbers of warnings which the master might have to take into consideration and their reliability is effectively overcome.

115. In a case such as the present, the master might fairly have obtained two authoritative forecasts not including warnings but there may have been another, which the master did not think as being as authoritative, which by chance had included an appropriate warning. This was not so in the present case but one may envisage many different situations in the South China Sea and elsewhere in the region. The evidence showed that there were a considerable number of reporting agencies whose broadcasts could be heard but which could not be understood by the master. Even if the words "in English" or "in the language customarily spoken by the master" were inserted, and that would certainly involve rewriting the condition, there would remain difficulties of selection and drastic consequences would follow from any failure to listen to every available transmission of that kind immediately before leaving harbour. If, however, knowledge is required before the master "allows" the ship to sail, those problems are overcome. I cannot therefore accept the arguments put forward in this respect by the appellants.

118. It would follow that the appellants have not made out the argument as to the construction of the policy relating to the knowledge of the master. There remains the subsidiary question whether the learned judge was correct in making the findings he did as to that knowledge.

121. The criticisms that the appellants made of the findings of the learned judge as to knowledge originally covered three distinct aspects. First, it was contended that the managing agent, Northern Star through Mr Shan, had sufficient knowledge of the warnings such that it was in a position to prevent the ship from sailing. Secondly, they said there was sufficient evidence of Mr Ruan's passing on the telex or at least some other form of the warning about the storm before the vessel left Beihai. Thirdly, they said that an examination of the telexes between the master and the respondent's agents contained sufficient in them to show that the master was aware of the storm warnings.

124. As to the argument that Mr Shan and Northern Star in Hong Kong had enough knowledge of these storm warnings to prevent the ship leaving port, it is sufficient now to say that counsel for the appellants explicitly abandoned reliance on the knowledge of Mr Shan for this purpose during the course of argument. They were right to do so for, although Mr Shan had sufficient knowledge and premonition to prompt him to send a warning telex, he did not, as pointed out by the respondent, have sufficient detailed knowledge of the storm warning and more particularly of the proposed movements of the vessel. At the relevant times he was not aware precisely when the loading of the clay would be finished or when the vessel could depart, nor had he been told of the proposed line of voyage or estimated time of arrival of the vessel. In those circumstances it is not necessary to examine further whether the knowledge had to be directly that of the master, or some person in similar control of the vessel, or whether a managing agent such as Northern Star could be said also to have been in a position to allow the vessel to sail or put out of port, within the meaning of the condition. There may be a good deal to be said for the view that, if knowledge is required, knowledge of an agent with authority to stop the vessel sailing would be sufficient, even if that agent did not pass on the information to the master. It is not necessary however to reach a conclusion on that aspect of the construction of the condition. There may have been similar difficulties in reaching a conclusion as to the knowledge of Mr Ruan, but his knowledge was not specifically relied upon on this appeal and it would seem that Mr Ruan himself and Penavico may not have had sufficient authority to prevent the master from sailing. Again, it is unnecessary to express a concluded view on this matter.

127. Insofar as they argued that the court might infer that Mr Ruan's attempt to communicate with the master was successful in the sense that thereby the master had knowledge of the storms, then the appellants faced a difficult task. In the first place the judge had explicitly held that Mr Ruan's own evidence was unreliable so that he could not make any findings based on that evidence. That was not surprising for the versions Mr Ruan gave of the events were both conflicting and confusing, and because examination revealed him to be more than unsure of the events which took place seven years earlier. For example, although he conceded he was aware of certain weather forecasts put out by the local meteorological station, he suggested in evidence that he had passed that on to the master at a time before they had been transmitted. Again, in attempting to justify his version of events that he handed the telex either to the master or, possibly, to somebody aboard the ship before it departed, he was led into saying that he did so during the morning before the inspection was completed. That entirely overlooked the fact that the telex was not transmitted until 12.22 by which time the inspection was completed and the ship had attempted to leave port but in fact had been delayed in the harbour, though not at the wharf, while it made various repairs to its engine. He was unable to explain how he then got the message to the master or anybody else on board.

130. As to the contention that an inference could otherwise be drawn notwithstanding Mr Ruan's evidence was not accepted, the argument largely rested, as I would understand it, on the fact that it had been received by 12.24 which was at least five hours before the vessel's actual departure from the harbour. I can see no error in the learned judge's finding that the telex was not delivered; nor do I think that he should have drawn an inference merely because of the passing of that length of time before the actual departure by the vessel. As I have said in the circumstances direct contact with the shore normally ended when an exit inspection was completed and a ship was cleared for departure. In this case the ship remained longer but not at the wharf, while repairs were being effected. I would not draw any inference against the respondent in this respect.

133. Finally, some time was taken in analysing a series of telexes sent by and to the ship after it left port. A series of some five telexes starting at 0812 and finishing at 1359 on 15 August were examined for the purpose of showing that one should draw an inference that the master had become aware of the storm warnings at the relevant time, i.e. before the vessel left port. Nothing in those telexes revealed that the master was aware of the warnings. Indeed, his own telexes or radio messages indicated the contrary for he seemed perfectly content to wait at the entrance to Hainan Strait for clearance and there was no suggestion that he was about to take evasive action as might otherwise have been expected. There was one message at 1131 from Northern Star to the master which contained the advice "take/extra caustions [sic] typhoon", but that was clearly sent after the vessel had left port. I am not able to draw any inference from this material against the respondent.

136. I am therefore not satisfied that the appellants have established that the master or other person in authority of this ship had knowledge of the relevant warnings at the time the vessel was allowed to leave the port of Beihai. The learned judge was correct in holding that it had not been proved that the master had such knowledge and, having regard to his construction of the clause which I have already stated to be the correct one in that respect, the appellants' contentions relating to knowledge have not been made out.

Whether "warning" included that of future courses of storm

140. It is desirable to return to the first ground argued on behalf of the appellants, namely that the warnings to which the clause refers comprehend more than warnings as to the present effects of a storm but also include warnings as to the future course of a storm in so far as it might affect a proposed voyage. The conclusion reached as to knowledge means, however, that even if the appellants be correct in this contention, it cannot affect the outcome of the appeal in that, on the proper construction of the clause, the master (or any other person in authority) was not shown to have had knowledge of the storm, either its present position or future course, and so the appellants have been unable to show that the vessel had been allowed to leave port under the proscribed conditions.

143. It would seem that reasoning based on the difficulties in giving the clause a precise meaning also took the learned trial judge to the conclusion that warnings of the kind referred to in the condition must be confined to those of an existing danger at the relevant places. He acknowledged that there were still some difficulties but, whatever they were, at the time of sailing, between 1730 and 1800 hours on 14 August, the severe storm was still two hundred or so miles away from any relevant point on the line of voyage here contemplated. I must say, however, that the conclusion had a very peculiar consequence inasmuch as such a limited interpretation would seem, on the detailed meteorological and navigational evidence here given, to have been virtually irrelevant so far as the port of destination was concerned. In the present case that port, being Hong Kong, was seen as close on three days' sailing away, yet the policy could be made void if there were on 14 August a storm warning then and there affecting Hong Kong and its anchorage. Bearing in mind the speed at which storms of this severity move, as the evidence showed, and also the likelihood of their dissipating after hitting land, there would seem little point in prohibiting the vessel from sailing if the storm was already directly over or affecting Hong Kong and thus likely to have moved on some time before the vessel approached Hong Kong. Indeed, for this reason, I first thought that the learned judge had also substituted "departure" for "destination" in the first part of the description of the area affected by the storm warning. That, at least, would have made sense, albeit that a master was hardly likely to set sail into the teeth of a storm of that magnitude.[6]

146. Be that as it may, I see considerable difficulties in accepting the learned judge's interpretation of the clause. Essentially, it is a clause about warnings, warnings of a particular kind, but warnings which are familiar to masters of vessels around the world and those engaged in merchant shipping. "Warning" is a word which connotes futurity. To the extent that it refers to an existing danger, it does so primarily because of potential misfortune in the future but it is primarily a word which connotes advice to take care against what may happen in the future. The cognate verb's primary meaning in the Oxford English Dictionary (2nd ed.) is "to give timely notice to (a person) of impending danger or misfortune". Indeed, in the case of meteorological warnings, their primary value is that they give forewarning of dangerous changes in the weather, such as storms, frosts, snow, floods and the like, so as to enable precautionary steps to be taken to avoid the harm of which warning is given. It may be said that there is relatively little point in giving warning of an existing storm or flood to those subject to them who well and truly know their effects without being told of them on a weather forecast which they would then have little time to take in.

149. Thus, in my opinion, weather forecasts and warnings of storms and other meteorological disasters ordinarily look to the future, providing warning of what may happen a day, two days or even further ahead. The universality of these considerations appears from the definitions appearing in the relevant manual of the World Meteorological Organisation, already set out[7] and apparently accepted by the learned judge. In each of the definitions of storm warning and typhoon warning the critical words defining the relevant message are that which is "intended to warn those concerned of the occurrence or expected occurrence of ..." the high wind speeds forecast. If, as seems highly likely, the reference in the condition to "typhoon or storm warning" was intended to incorporate that understanding of the expressions, then the understanding included not merely the existing occurrence of high wind speeds but, more especially, of the "expected occurrence" of high wind speeds likely to result in high danger to the vessel. Moreover, the condition should be read in the light of the kind of forecast warnings customarily put out and relied upon in the South China Sea, namely those of ROHK and JMA. They each gave an existing position and then, as part of the warning, proceeded to give a direction and speed of movement of the storm together with estimated positions thereof 24 and 48 hours in the future. There was much argument on the part of the respondent that the further one went away from the existing centre of the storm and the longer in the future the forecast was intended to cover, the more imprecise was the area affected by the warning, such that it was not fair or reasonable to require the warning to be accepted by the insured and made it unlikely the master would place reliance on it. There are some difficulties, as will be seen below, but coverage of a period of 48 hours seems not unreasonable to expect in warnings of this kind. One is aware that certain forecasters and bureaux attempt to give some form of weather forecast up to seven days into the future, but the present warnings were limited, fairly one might say, to 48 hours and that is a period which might reasonably be taken into account by the master of a vessel when setting sail, even if it be only for his own protection and that of the crew. The condition still must be interpreted to require some coincidence of the vessel's presence and the arrival of the storm.

152. The potential difficulties arising from a master's possible ignorance of particular bureaux and particular warnings have already been dealt with but there remains one issue arising out of the existence of a number of bureaux and several warnings being given (or not given) in relation to the same proposed voyage. It was pointed out that even in the present case the forecast course of the storm varied between the two agencies. So it was said that it might be possible for one bureau to give a forecast warning which did not cross the proposed course of the vessel, while another would affect the voyage. Furthermore, it was observed that the Biyayang Ginto went down in an area not strictly covered by either of the forecasts, because the typhoon veered a little to the south-west after first crossing land.

155. To my way of thinking these objections are unrealistic and do not give sufficient commercial efficacy to the clause. Warnings of the relevant kind must, of course, refer only to risks of storms capable of causing a disaster to a subject vessel. The condition would seek to take off cover a vessel whose master chose to put to sea in the face of warnings of this kind. One would think as a matter of good seamanship that a master would not choose to run the risk of a severe storm or typhoon even if one (or even more than one) of the storm warnings did not forecast the storm as crossing the course which the vessel proposed to take. The master is not intended to be an expert meteorologist but is supposed to have sufficient commonsense to know that when such storms arise there is a real risk that they will cross the path of the vessel (in general terms) and place the vessel at risk of sinking or suffering severe damage. No competent masters would choose to take such foolhardy risks merely because they thought that one forecast or the other was more reliable. The object of the condition is to prevent such risks being taken and in that sense the condition can be given a perfectly sensible construction by requiring the master not to allow the vessel to sail even if only one warning of several would place the storm in the path of the proposed voyage.

158. The same factors ought to be seen to prevail where issues of certainty have been raised as to how one should precisely identify the position of the storm and its forecast course. Many of the difficulties raised by the respondent apply equally to the interpretation placed on the condition by the learned judge, which the respondent supports, as well as to the contentions of the appellant. Thus, it is acknowledged always that there are difficulties in identifying where a storm is at any particular moment; the eye of a typhoon is ordinarily calm, but dangerous winds extend around that eye for several nautical miles. So, on the respondent's contentions, one would still have a difficulty if the precise eye of an existing storm had been, for the purposes of the present voyage, say 20 miles east of Hong Kong on 14 August. Likewise, there might be difficulties if a typhoon had been travelling parallel to the proposed line of voyage but at the relevant time was never closer than 20 or 30 miles to that proposed course, as in fact occurred in the later stages of typhoon Fred's path towards the Luichow Peninsula, before it crossed and turned due west shortly before the sinking of the Biyayang Ginto. The difficulty would exist still, perhaps to a lesser degree, if the respondent's interpretation as to an existing storm had to be applied only to a proposed voyage. Moreover, to that element of uncertainty must always be added that of defining the proposed line of voyage of any vessel from one port to another, except in the case where the vessel is travelling through, say, a strait or is required to follow a precise course. In the present case the vessel was marked on the exhibited charts as likely to take a course directly from the exit of Hainan Strait to Hong Kong, but the master was not obliged to do so and, as was pointed out, he could have changed course, at least to some extent, to avoid the worst of any existing or forecast storm.

161. Do these factors, however, face the reality of what this condition is attempting to achieve? It is the risks of disaster which are to be avoided so that the ship is not to be allowed to set sail if those risks can fairly be said to exist. If that be the proper approach, then its interpretation should not confine the risk to a storm at a precise point on a meteorological map, even if one could identify that at the moment before sailing. Thus, the position of the storm, even at the outset, must be viewed as a circle based on the stated position. It will be noted that the ROHK warnings were expressed in terms that the severe tropical storm was "centred within 60 nautical miles" of the stated point. Further, the same warning identified the radius of severe tropical storm winds, i.e. those over 47 knots, as extending for the same 60 nautical miles. The JMA warning was not expressed in exactly the same terms inasmuch as it identified a precise position, although the maximum winds were described at the relevant time as being "55 knots near centre". However, it also contained statements as to the strength of winds within two specified radii, namely that 50 knot or greater winds were to be found being within a radius of 30 miles and 30 knot or greater winds within a radius of 130 miles, each at 1400 hours. The relevant radius would therefore seem to be that applicable to winds over 50 knots, being in that case 30 miles. Taking that broad but practical view of these warnings, the difficulties would be the same at the fringes of the relevant area whether or not the respondent's or the appellant's contentions were accepted, but they cannot prevent the condition from being given a sensible construction.

164. There is, it must be conceded, a further difficulty upon which the learned judge and the respondent placed heavy weight in order to show that future warnings were not covered. That arises because the forecast path of any storm must have a further element of uncertainty in it. Of course, both the ROHK and JMA warnings gave precise positions, in a sense, for both their 24 and 48 hour warnings. The ROHK warning did not refer to the possibility of any deviation but the JMA warning did, in the case of the 24 hour warnings being expressed to have an "uncertainty of 90 miles radius" and, in the case of the 48 hour forecast position, expressed in terms of an "uncertainty of 180 miles radius". This was described as creating a "cone" effect or an area of expected occurrence which would spread wider, as time passed, in a way which could only be mapped in the form of a cone. This was said by the respondent to create too much uncertainty such as would affect a voyage which might pass temporarily through such a zone although the voyage might be in a "different direction" and therefore, because of timing and other factors, might be realistically outside any real risk. For the purpose of this condition, however, the forecast warnings must refer to mapped positions, the element of uncertainty being merely an indication to mariners that forecasting of this kind has its necessary element of uncertainty. For practical purposes, however, the path of the storm should be taken as an area or zone following the forecast mapped positions and extending on each side 20 or 30 miles according to the radius then existing (or forecast) of winds which satisfy the tests for severe storms or typhoons. Although that may not be necessarily a straight zone, the area affected could be understood and worked out by a master at the relevant time with reasonable certainty, upon which he might be expected to act. It should be remembered that, in the end, the master will not be so much influenced by the condition in the policy as by the practical dictates of seamanship and an appreciation of risks that a master would not ordinarily run as a matter of commonsense.

167. On that basis I can see no real practical difficulty in interpreting the condition upon the basis that it applies where the zone of the storm or typhoon of which warning is given will affect in the future the port of destination (in this case Hong Kong) or the line of voyage between the port of departure and the port of destination, in this case the proposed journey from Beihai south and through the Hainan Strait and thence north-east to Hong Kong. Whatever else the vessel might do to avoid such a storm or typhoon, the assumption is that, when a ship sails, it proposes to sail to a specific port of destination: if it does not so sail, the clause will not apply unless a relevant warning affects some new voyage or destination. In the present case, there can be no doubt that the Biyayang Ginto took the course it originally proposed, south to the entrance to Hainan Strait, waiting there for a passage through the strait and thence directly to Hong Kong. This much is clear from the master's last message (before his SOS) when at 1359 on 15 August he stated his estimated time of arrival at Hong Kong to be 1500 hours on 17 August.

170. The final, and in many senses the most difficult, question arising in relation to this aspect of the appeal concerns timing, both in the sense of estimated times of departure and arrival for the subject vessel where the condition may operate to affect a particular voyage and secondly, as to the times at which the storm or typhoon might be expected to affect that voyage, both its destination and also the line of voyage to be taken by the vessel. It should be remembered in each case that the decision taken to allow the ship to sail must be made before it leaves port, for the proviso, if it be properly so described, makes clear that the warranty no longer applies "should the vessel have sailed out of port prior to such warning". In practical terms, therefore, in the present case the last ROHK warning fixed a position for the severe storm about an hour (possibly as short as half an hour) before the ship's departure and in the case of JMA the last warning was about four hours before its departure.[8]

173. In a manner of speaking, therefore, the warnings were, as present or existing warnings, already out of date by the time they could fairly have reached the master of the Biyayang Ginto. More importantly, the same difficulty would arise, to a greater or lesser extent, each time an insured vessel left port. As has already been mentioned, the relevance of a warning at the time of leaving port then and there affecting the port of destination must be seen to be remote. It was argued that many voyages might be shorter than the proposed voyage from Beihai to Hong Kong, so that the port of destination might have been reached rather earlier than the estimated time of arrival which in this case was close to three days after departure. While I acknowledge that this may be so, the practical likelihood of there being many voyages involving the carriage of cargo for short distances would be relatively small, unless the vessel voyaged further south between the islands of the Philippines or Indonesia. On the other hand, one may well envisage voyages both of this length and of much greater length through these seas which might involve this relatively slow-moving vessel in being at sea for several days. Why therefore those responsible for drafting the clause, if it contemplated only existing storm warnings, i.e., warnings of existing storms, would first refer to a typhoon or storm warning at the port of destination is hard to fathom. It would be highly unlikely that a storm or typhoon of the relevant kind would stay at that one position for so many days. In this particular way the language of the condition points to the unlikelihood that the word "warning" was intended to convey only warnings of the presence in the stated positions of existing storms or typhoons.

176. It is possible, however, to envisage timing difficulties which the parties may not have contemplated. A 48 hour warning, on the appellant's contentions, may be forecast to affect the first few miles of a proposed voyage many hours after the ship had sailed through those waters and was at that time many hundreds of miles away. Likewise, one can see in a longer voyage the possibility of a storm being forecast to affect the line of voyage or port of destination a day or more before the vessel could possibly arrive in those waters. Nevertheless, the condition is concerned with risks, risks of a serious kind, and it would not be inappropriate to apply a broad brush approach to the practical application of this condition having regard to the known nature and form of storm warnings. In practical terms, at least so far as the evidence in the present case goes, future warnings extend only to a period 48 hours from a time before the vessel departs and possibly from a time several hours before it departs. The kind of storm covered by the clause moves at considerable speed, indeed, at a speed often far faster than that which can be attained by the kind of vessels covered by this policy. There will always be doubt as to precisely where the storm will go and when it will arrive at any point on a predicted path. But the dictates of good seamanship would suggest that they are the kind of risks which a master would not run by sailing into an area covered by a forecast storm. The respondent put considerable weight on the fact that the ship went down at a point not on the path which was forecast by either bureau, indeed at a point which was never forecast to be crossed by the storm. That, however, only points to the fact that these are severe risks which the parties to the policy agreed ought not to be run by the master or any other person responsible for allowing the vessel to sail. If voyages ordinarily take a day or so to complete, so also should the relevant storms cover a period of 48 hours or, in the case of a shorter voyage, possibly only a period of 24 hours. The primary difficulty arises in cases of voyages which take longer and where the risk is not forecast to cover the line of voyage at a time likely to coincide with the passing of the vessel along that line.

179. For this purpose I do not believe the condition should be read as requiring any precise estimate to be made of the point which the ship might have reached at the relevant time the storm was forecast to cross that line. So long as the two paths were likely to cross at some time during the voyage, then the condition would be satisfied if the vessel were allowed to sail. Of course one must pay regard to the limit of the forecast warning, whether that be 24 or 48 hours. If a voyage is likely to be longer than 48 hours then the possibility of a storm affecting the later path of the voyage after the expiry of the 48 hours might fairly be disregarded. In the present case each of the two warnings forecast severe storms as crossing the direct line of the voyage of the vessel within the 48 hour period. If there were knowledge of either warning the vessel should not have been allowed to sail.

182. Finally, I turn to an argument accepted by the learned trial judge and maintained on this appeal by the respondent to the effect that the reference to the typhoon or storm warning in the condition is expressed in the present tense. As argued by the respondent the subject of the verb "is" is the words "typhoon or storm", so that the relevant warning must be of a typhoon or storm which presently is "between the said port" or is at the port of destination. This seems to be a misconception of the structure of the sentence. What the subject of the verb "is" is "a typhoon or storm warning". If, as the judge correctly pointed out, which was not disputed by the respondent, the definition of those terms from the World Meteorological Office comprehends both "the occurrence or expected occurrence" of the relevant conditions then, although there must be a "present" warning, that warning can, and in my opinion should, include warnings both of expected occurrences and conditions as well as of existing occurrences or conditions. The word "is" was used because the relevant clause refers to a time when the vessel is allowed to sail or put out of port and it "is" at that time that the relevant warning must exist. It says nothing as to whether the storm or typhoon then exists and the term "warning" most naturally comprehends future events as much as any present event likely to cause risk of damage.

185. It is for those reasons I cannot, with great respect, accept the construction of this aspect of the clause by the learned trial judge and must reject the contentions of the respondent in that regard. If it were not for the matter of knowledge, the first ground would have been made out as the evidence clearly indicated that at the relevant time there were warnings of a severe storm the path which would clearly cross that of the anticipated line of voyage of the Biyayang Ginto.

189. Although I would have held that the appellants should succeed as to their contention as to the nature of the warnings comprehended by the "typhoon clause", they have failed on the grounds relating to the master's alleged knowledge of the warnings, both as to interpretation and as to the factual findings. Their success on the first ground cannot overcome their inability to prove that the vessel was allowed to leave port with knowledge of the relevant warnings. Consequently the appeal must be dismissed so far as the principal issue of liability is concerned.

193. The issue as to interest was confined in the end to an argument that the learned judge had no discretion to add one per cent to the interest rate which was established by reference to US prime rates proved in evidence on the hearing. There is undoubtedly a wide discretion which still remains to a trial judge in awarding interest. The question is whether it was appropriate in the exercise of that discretion to add one per cent to the rates proved in evidence. It is argued that interest is essentially compensatory and not intended by way of penalty. That may be so in many jurisdictions but it is by no means clear in Victoria[9] where the applicable rate is fixed pursuant to the Penalty Interest Rates Act 1961 and has ordinarily been well in excess of prime rates and for some years has exceeded ordinary lending rates. It has been the effective practice over many years for the Attorney-General to prescribe an interest rate which has been in excess of most market rates.

196. Whatever be the correct view as to the extent of the power to award interest under ss.58 and 60 of the Supreme Court Act 1986, s.59 makes no reference which would incorporate the rates fixed under the Penalty Interest Rates Act 1983. Section 59(2), which it is agreed is the relevant section for present purposes, reads:

"The Court, on application in all proceedings on any policies of insurance, must, unless good cause is shown to the contrary, give damages in the nature of interest over and above the money receivable."

It would therefore appear that in relation to that sub-section at least the general principle often referred to in the High Court would apply, namely, that the purpose of such an award of interest is "to compensate the plaintiff for the detriment that he has suffered by being kept out of his money, and not to punish the defendant for having been dilatory in settling the plaintiff's claim"[10]. The section must therefore be viewed as founded on the need to compensate the plaintiff for being kept out of its money, so that prima facie there should be a factual basis for any award of interest thereunder. In most cases the dispute arises in relation to claims within the jurisdiction so that the Court may take judicial notice of or otherwise inform itself of lending and borrowing rates and related financial practices. Doubtless, this explains the addition of one per cent to the minimum bank lending rate by the English Commercial Court, as referred to in Miliangos v. George Frank Textiles Ltd. (No. 2)[11] and BP Exploration Co. v. Hunt (No. 2)[12]. However, whatever may be appropriate for determining matters of banking and financial practice within this jurisdiction, the Court cannot act in that way in relation to the award of interest at foreign rates in circumstances where the calculation must depend upon foreign banking and other financial practices.[13] In such cases this Court must act only on evidence, although it is not obliged to award the whole of the amount sought and justified by such evidence. The present case is one where there is evidence of a relevant foreign rate which the judge here accepted and was not challenged in this Court. There was no further evidence, as I would understand it, upon which the judge might act to increase the rate which is thereby said to be the measure of the respondent's loss.

201. Nevertheless, a discretion is given to a trial judge which must come within that which the statute permits. Where s.58 or s.60 applies, either the judge awards interest at the rate prescribed by statute or the judge chooses deliberately to impose a lower rate.[14] It has never been accepted that the judge can impose a higher rate unless that is done in accordance with contract[15] or specific evidence which would justify the imposition of that higher rate. Likewise, where the evidence as to an appropriate rate depends upon rates applicable in other jurisdictions, the judge ought ordinarily to apply the rate, or one of the rates, established by evidence, subject only in certain circumstances to a proper application of a judicial discretion to impose a lower rate. Again there would appear to be no basis for allowing a higher rate without there being some specific basis for it, whether based on evidence or upon the statutory regime for awarding interest.

204. A most careful argument presented by junior counsel for the appellants has set out in great detail the relevant authorities and considerations. It is sufficient to acknowledge the care with which that memorandum has been prepared, but it supports the argument originally put forward and not, in my opinion, satisfactorily answered by the respondent, that the discretion invested in a trial judge does not include the imposition of a higher rate unless there can be some specific basis for justifying that rate. Nothing in the learned judge's reasons for judgment would, I regret to say, justify the addition in this case, albeit that it is but one per cent. In the circumstances, I believe that the trial judge's discretion has miscarried in this respect and I would set aside the judgment for interest and direct that new calculations be

made upon the basis of the proved rate. If counsel are able to agree upon the figure then judgment should be given for the agreed sum. To that extent only I would allow the appeal.

209. In my opinion the "typhoon clause" required knowledge[16] of a typhoon or storm warning on the part of a person, in this case the master, who allowed the vessel to sail. That is the natural implication of the word "allowed" and overcomes most of the difficulties that would otherwise beset this aspect of the construction of the clause.[17] Accordingly the appeal as to liability fails on the facts, for the reasons given by Ormiston, J.A. at [37]-[42] and Batt, J.A. at [69]. I respectfully agree that there was no basis for increasing the rate of interest by one per cent and that the appeal should succeed on that issue.

214. I agree with Ormiston, J.A., but add some observations of my own on the two issues in the appeal that relate to the "typhoon clause".

217. In the interpretation of that clause the first and critical question is whether it required that the vessel should not, otherwise than under compulsion, sail or put out of port, if as a matter of fact there was a relevant typhoon or storm warning at the time of sailing or putting out, or whether, on the other hand, it meant that the insured was not to allow the vessel, when there was such a typhoon or storm warning, to sail or put out of port, upon the basis that the word "allow" connotes that the insured knew[18] of the warning that existed. In favour of the first

interpretation are simplicity and greater certainty of operation (leaving aside the definition of the warning), and the fact that it was clearly such an arrangement as the parties might have made. On this view what was not to be allowed was simply sailing or putting out of port, and "allow" merely referred to the voluntary nature of the sailing or putting out. The passive "shall not be allowed" may be said to show no concern with the agent or actor. But after reading and re-reading the clause I am satisfied that the complement of "shall not be allowed" is not the infinitives by themselves: with them must be read the temporal or "when" clause. In addition, someone has to allow a vessel to sail or put out, and in this context it had to be the insured (considered as an entity). It follows from these considerations that a vessel could not be "allowed" by the insured to "sail or put out when" there was a typhoon or storm warning unless the insured knew (or, perhaps, ought to have known) of the warning. I recognise that the meaning of "allow" in the present clause depends on its own context and is not determined by cases decided on different contexts. But the cases discussing the meaning of "allow", most of which have concerned prosecutions, do manifest a general requirement of knowledge[19]. There are suggestions in Hardcastle v. Bielby[20] and possibly in Gilbert v. Gulliver[21] that constructive knowledge may be sufficient, but Tanos[22] is to the contrary. The construction of the warranty as breached if there was constructive knowledge of a relevant warning on the part of the insured was very much an afterthought and secondary argument of the appellants and they did not advance any argument that the insured ought to have known through the medium of any other person than the master. The reasons of Ormiston, J.A. show why constructive knowledge cannot be sufficient in the case of the present clause.

221. The appeal was conducted before us on the basis, asserted by counsel for the respondent and not challenged by counsel for the appellant (if not expressly agreed in by him), that the onus of proof with regard to the typhoon clause, a warranty and condition, was upon the appellant insurers to prove non-satisfaction or breach of it.

224. It was only through the master, so the appellants argued before us, that the insured knew of the warnings. An argument that Mr. Shan, the manager of the agent, knew was abandoned at the opening of the appeal.

227. As the first step in an argument that knowledge on the part of the master should be inferred, it was argued that we should infer that, as requested in Northern Star's telex of 1222 hours on 14 August 1991, which was marked "Urgent Urgent" and was received more than five hours before the vessel sailed, Mr. Ruan delivered to the captain the message contained in the telex. I am quite unpersuaded to draw this first requisite inference: the evidence of Ruan having been entirely rejected and that rejection being accepted, it is not possible, in my view, to infer that the message reached the master, otherwise than by Ruan, in the ordinary course of events, for, as the respondent's counsel said, the difficulty is that, on the evidence, the ordinary course of events involved, on the limited evidence available, delivery by Ruan. Moreover, the fact that the vessel sailed at the least suggests that the master did not receive the message in the telex. Since the appellants' argument fails at the first inference, it is unnecessary and indeed not possible to consider whether the second inference (that of knowledge of a relevant warning) should be drawn.

230. Since knowledge on the part of the insured of neither warning was proved, it is not really necessary otherwise to pass upon the construction of the warranty. However, I should say that I cannot accept his Honour's construction of it as referring, and indeed referring only, to a warning current[23] at the time of departure of the then present existence of a storm or typhoon at the point of destination or at any point along the then intended course of the vessel to the port of destination. This interpretation ignores the fact that the more frequent use of the word "warning" is to draw attention to an event or state which is likely to occur or come into existence rather than one which exists or is occurring. It also ignores the reference to "expected occurrence" in the accepted definitions of storm warning and typhoon warning. Nor is it supported by the present tense "is" in the temporal clause in the condition, for that is the time of existence of the warning itself and the clause says nothing about whether the present warning is as to a present occurrence or state of affairs or as to a future occurrence or state of affairs. Secondly, his Honour's interpretation leads to results which are not only inconvenient but irrational and not such as commercial persons might have had in mind. For example, according to it the insured could without breach of warranty allow the vessel to sail into the face of an expected typhoon simply because the typhoon was not at the time of sailing located on the intended course of the vessel, even though it was predicted to cross within a few hours that course at a point where the vessel was expected to be then.

[2] It was accepted on both sides, and I am content also to accept, that the onus rested on the appellant insurers to show that the terms of the condition had been satisfied.

[3] See, e.g., Australian Casualty Co. Ltd. v. Federico [1986] HCA 32; (1986) 160 C.L.R. 513; Johnson v. American Home Assurance Co. [1998] HCA 14; (1998) 192 C.L.R. 266.

[5] The second ground in fact argued: see [24].

[6] Nevertheless some of the expert evidence suggested that remaining at anchor in certain harbours was in fact more dangerous than leaving port and that ships were sometimes advised for that reason to sail out of them as quickly as possible.

[8] This, of course, makes no allowance for the fact, probably irrelevant to the clause's interpretation, that it was not possible to use the radio room immediately before departure from Beihai so that any warnings had to be taken to the ship or transmitted on VHF radio from shore to ship at some time which suited the convenience of the local agent Penavico.

[9] See the comment by the majority (Gleeson, C.J., Gaudron, McHugh, Gummow, Hayne, JJ.) in Grincelis v. House [2000] H.C.A. 42; 74 A.L.J.R. 1247 at 1251 para.[16].

[10] See Batchelor v. Burke [1981] HCA 30; (1981) 148 C.L.R. 448 at 455 per Gibbs, C.J.; Haines v. Bendall [1991] HCA 15; (1991) 172 C.L.R. 60; MBP (SA) Pty. Ltd. v. Gogic [1991] HCA 3; (1991) 171 C.L.R. 657 at 663; Grincelis v. House [2000] H.C.A. 42; 74 A.L.J.R. 1247 at 1251 para.[16]; State Bank of N.S.W. v. Swiss Bank Corp. (1995) 39 N.S.W.L.R. 350 at 360 (C.A.).

[11] [1977] Q.B. 489 at 495.

[12] [1979] 1 W.L.R. 783 at 849.

[14] See ss.58 and 60 of the Supreme Court Act 1986 ("not exceeding the rate"): cf. Clarke v. Foodland Stores Pty. Ltd. [1993] VicRp 81; [1993] 2 V.R. 382 at 394, 396-397.

[15] See s.57(1) of the Supreme Court Act.

[16] Wilful blindness would almost certainly be equivalent to knowledge for this purpose: cf. R. v. Crabbe [1985] HCA 22; (1985) 156 C.L.R. 464 at 470-471 and Manifest Shipping Co. Ltd. v. Uni-Polaris Shipping Co. Ltd. [2001] UKHL 1; [2001] 2 W.L.R. 170 especially at [26].

[17] I do not wish to say anything else about the construction of the clause except to express my concurrence in Batt, J.A.'s observations at [70].

[18] Or, perhaps, either knew or ought to have known.

[19] Massey v. Morriss [1894] 2 Q.B. 412; Crabtree v. Fern Spinning Co (1901) 85 L.T. 549; Gilbert v. Gulliver [1918] VicLawRp 27; [1918] V.L.R. 185; Victor v. Chief of Naval Staff (1993) 115 A.L.R. 716 at 723; and Commissioner of Police v. Tanos [1958] HCA 6; (1958) 98 C.L.R. 383 at 394.

[20] [1892] 1 Q.B. 709 at 712-713.

[23] A recent warning that has not been cancelled or withdrawn would ordinarily qualify as a warning current at the time of departure.