Supreme Court Judgments

Supreme Court of Canada

Sherwin-Williams v. Boiler Inspection, [1950] S.C.R. 187

Date: 1949-12-22

The Sherwin-Williams Company Of Canada Limited (Plaintiff) Appellant;

and

Boiler Inspection And Insurance Company Of Canada (Defendant) Respondent.

1949: May 13, 16, 17, 18; 1949: December 22.

Present: Rinfret C.J. and Taschereau, Band, Estey and Locke JJ.

ON APPEAL FROM THE COURT OF KING'S BENCH, APPEAL SIDE, PROVINCE OF QUEBEC

Insurance—Against damage caused by accident—Policy excludes loss from fire and from accident caused by fire—Accident followed by fire and explosion—Whether loss covered—Cause of—Assignment of insured's rights—No signification—Whether insured can still claim—Arts. 1570, 1571 C.C.

An insurance policy insured appellant against loss on property directly damaged by accident and excluded losses from fire and from accident caused by fire. A tank, which was the object of the insurance, burst permitting the escape of fumes which ignited and exploded causing considerable damage to appellant's factory. The Superior Court maintained the action on the policy and the Court of Appeal dismissed it on the ground that the damages were caused by fire and were not the direct result of the tearing asunder of the tank.

[Page 188]

Held: The damage was the direct consequence of the accident to the tank; the bursting of the tank was the proximate cause of the damage. Coxe v. Employers' Liability Ass. Corp. (1916) 2 K.B. 629; Leyland Shipping Co. v. Norwich Union Fire Ins. Society [1918] A.C. 350 and Canada Rice Mills v. Union Marine and General Ins. Co. [1941] A.C. 55 referred to. Stanley v. Western Ins. Co. (1868) L.R. 3 Ex. 71 distinguished.

Held also, that the appellant was not deprived of its right of action against the respondent, as the assignment of its rights to the fire insurance companies had not been signified to the respondent.

Per Band (dissenting): The explosion damage was attributable to the fire which, existing briefly after the initial stages of the accident to the tank, caused the explosion and was a new point of departure in the chain of causation.

APPEAL from the decision of the Court of King's Bench, appeal side, province of Quebec 1, reversing the judgment of the Superior Court, Tyndale CJ., and dismissing appellant's action on an insurance policy.

J. A. Mann, K.C. for the appellant.

John T. Hackett, K.C. and L. P. Gagnon, K.C. for the respondent.

The judgment of the Chief Justice and of Taschereau J. was delivered by

Taschereau J.:—The first point that has to be dealt with, is the question of the appellant's interest. It is contended on behalf of the respondent that the appellant shortly after the institution of the present action, having transferred and assigned to the fire insurance companies, all its rights against the respondent, for and in consideration of the sum of $46,931.28, cannot succeed for lack of interest.

With this proposition, I do not agree, as I think that even if the appellant had assigned its rights before the action was started, without the necessary signification being given, it would still have the necessary interest to claim from the respondent.

The assignees of the claim did not insure the appellant assignor for damage caused by accident. Their policies covered damage caused by fire, and in this respect they have fulfilled their obligation, by paying to the appellant the full amount of its losses. But they have additionally paid

[Page 189]

$46,931.28 for the damage caused by an explosion, which the appellant now says is covered by the respondent's policy. Assuming therefore the liability of the defendant, it necessarily follows that the fire insurance companies are not the appellant's insurers for the damage now claimed in the present action.

We are not confronted here with the case of an insurance company which, after having paid its own client, victim of an accident, the amount to which the latter is contractually entitled, obtains a subrogation receipt against the tort-feasor. In such a case, there is no doubt that the victim, although having signed a subrogation receipt, may still claim against the author of the damage he has suffered. The legal relations that exist between the victim and the insurer are obviously contractual; those between the victim and the wrongdoer are delictual. They are two entirely different causes of action. It is for his own protection that the victim has paid to obtain compensation, and not for the benefit of the wrongdoer. The latter has no concern with the rights of the insured and the insurance company inter se.

In such a case the rights of the victim to sue the author of the tort have been often recognized. Vide (McFee & Co. v. Montreal Transportation Co. 2); (Millard v. Toronto R.W. Co. 3).

In Hebert v. Rose 4, the Court of Appeal of the Province of Quebec held:—

Where a certain sum is found to be due for damages caused to an automobile through a collision, an amount received by the plaintiff from an insurance company which had insured his automobile against loss or damage through collision, cannot be deducted from the award.

And later in Coderre v. Douville 5, Mr. Justice Rivard, speaking for the same Court, said:—

L'appelant va plus loin; il soutient que de demandeur n'a pas le droit aux dommages-intérêts parce qu'il a déjà été indemnisé par la compagnie d'assurance, qu'il y aurait eu subrogation et novation. Les termes de l'acte intervenu entre le demandeur et l'assureur sont clairs; c'est bien une cession de ses droits que Douville a consenti. Dans ce cas, le recours au nom du créancier contre l'auteur du dommage reste ouvert.

In all these cases, the plaintiffs had been paid by their insurers, but this jurisprudence cannot determine the rights of the plaintiff in the case at bar. I have referred to it,

[Page 190]

merely to point out the entirely different rights of the plaintiff, and to avoid any further confusion on the matter. It may also be said that the amendment to section 2468 C.C. enacted by the Quebec Legislature in 1942 (6 Geo. VI, chap. 68), which says that civil responsibility shall in no way be lessened or altered by the effect of insurance contracts, would cover cases similar to those which I have cited. The object of this section being to confirm the principle established by the Court of Appeal of Quebec, that a wrongdoer may not deduct from the amount of damage he has occasioned, the moneys received by the victim from an insurance company.

In the present case, the various fire insurance companies, the transferees of the claim against the respondent, are not insurers against damage originally caused by explosion. They are assignees of a debt which they have bought from the appellant, and therefore, different principles have to be applied.

The two relevant sections of the Civil Code are 1570 and 1571. They read as follows:—

1570. The sale of debts and rights of action against third persons, is perfected between the seller and buyer by the completion of the title, if authentic, or the delivery, of it, if under private signature.

1571. The buyer has no possession available against third persons, until signification of the act of sale has been made, and a copy of it delivered to the debtor. He may, however, be put in possession by the acceptance of the transfer by the debtor, subject to the special provisions contained in article 2127.

Between the appellant and the fire insurance companies, the sale was perfected at the date the relevant document was signed, but it is not contested that a copy of it has never been delivered to the respondent. Of course, this was essential to give the insurance companies "possession available" against the respondent, but it is argued that although the assignees could not exercise their rights until the fulfilment of this requirement of the law, the assignor was nevertheless divested of all his rights of ownership, and could not properly bring the present action. If he did so, it would be in violation of section 81 of the Code of Civil Procedure, which says:—

81. A person cannot use the name of another to plead, except the Crown through its (recognized officers.

[Page 191]

It has been said that this theory has received the support of Mr. Justice Cimon in Montreal Loan & Investment Co. v. Plourde 6. But I do not think that such is the case. A perusal of that judgment shows that the plaintiff, the assignor had sold to the assignee a claim against the defendant, but the latter, in lieu of notification, had accepted the assignment. The learned judge rightly decided that the assignee was the only proper party who could claim, haying, on account of the acceptation by the debtor, a "possession available" against him. In view of 1571, the assignor was divested of all his rights, and any action taken by him was in the name of "another" and contrary to 81 Code C.P.

But here, there was no notification, no acceptance, and if, between the seller and the buyer the deed of sale was complete, it was not as to third parties. Until the signification is made, as to third parties, the title remains in the assignor. This is so true, that a garnishee may be served in execution of a judgment against the assignor upon moneys in the hands of the debtor. The former or the assignee will not be allowed to oppose the transfer, if no signification has been made. Vide Aubry & Rau (Traité Pratique de Droit Civil, Vol. 7, p. 450).

Article 1690 of the French Civil Code is similar to section 1571 of the Quebec Code. The French authors are unanimous to accept the theory that until a copy of the deed is served upon the debtor, the title, as to third parties, remains vested in the assignor, who alone may properly bring action to recover the debt.

Troplong (Droit Civil Français, De la Vente, Vol. 2, 1854, page 457) says:—

Si la signification eat encore à faire, le cédant poursuivra le débiteur sans que celui-ci puisse lui opposer que, lui cédant, il s'est dépouillé de ses droits. C'est ce qui été jugé par arrêt de Ja Cour de Cassation du 4 décembre 1827, portant cassation d'un arrêt de la Cour de Calmer, du 27 août 1824.

Zachariae (Le Droit Civil Français, Vol. 4, 1858, pp. 326 and 327) expresses his views as follows:—

Il y a plus, tant que le cessionnaire n'est pas saisi, le cédant lui-même peut exiger le paiement sans que le débiteur cédé puisse lui opposer la cession qu'il en a faite.

[Page 192]

Aubry & Rau (Droit Civil Français, tome 4, 4th Ed. 1871, p. 434) share the same opinion:—

Quant au cédant, il conserve jusqu'à la signification ou acceptation du transport, le droit de faire, tant à l'égard des tiers qu'à l'égard du débiteur, tous les actes conservatoires de la créance, et même celui d'exercer les actions et poursuites y relatives.

Planiol & Ripert (Traité Pratique de Droit Civil Français, Vol. 7, 1931, p. 449) also says:—

Pendant l'intervalle qui sépare la cession de l'accomplissement de l'une. des formalités de l'article 1690, la créance qui appartient déjà au cessionnaire à l'égard du cédant, appartient toujours au cédant au regard des tiers.

At pages 449 and 450, the same author says:—

Ainsi jusqu'à l'acceptation ou à la signification, le cédant peut poursuivre le débiteur ou recevoir paiement.

And on the same page:—

Une fois la signification ou l'acceptation intervenue, la situation est renversée. Le cédant est sorti du rapport d'obligation à l'égard de qui que ce soit, le cessionnaire seul se trouve investi de la qualité de créancier.

Laurent (Principes de Droit Civil Français, Vol. 24, 3rd Ed. pp. 499 and 500) teaches that:—

L'article 1690 porte que le cessionnaire n'est saisi à l'égard des tiers que par la signification du transport, ou par l'acceptation que le débiteur en a faite dans un acte authentique. De là suit que le cédant reste saisi de la créance à l'égard des tiers, malgré le transport qu'il en a fait, jusqu'à ce que la cession ait été signifiée ou acceptée. C'est ce que dit Pothier; et quand il dit que le cédant n'a point été saisi de la créance, cela signifie qu'il en reste propriétaire.

On the same page:—

Le cédant reste propriétaire de la créance à l'égard des tiers, le débiteur est un tiers; donc, le cédant reste créancier et le débiteur est tenu de payer, et il a aussi le droit de payer.

To the opinion of these learned authors may also be added, what Mr. Justice Rinfret, now C.J., said in the case of Lamy v. Rouleau 7. Although section 2127 C.C. was there invoked, which is not the case here, there are some principles which have been enunciated in that judgment, which are useful in the determination of the case at bar.

The assignor, remaining the creditor cannot be considered as claiming in the name of another, in violation of section 81 of the Code of Civil Procedure. He has the right to sue in his own name, because as to third parties, the title is still vested in him.

[Page 193]

It is possible that a different situation would arise if even before a formal signification, the assignee instituted proceedings to recover the amount due by the debtor, because in such a case his action would in itself be a sufficient signification of the act of sale, as decided by the Judicial Committee of the Privy Council in Bank of Toronto v. St. Lawrence Fire Insurance Co. 8, but this is not the case here.

I have therefore to come to the conclusion that the plain- tiff had a sufficient interest to institute the proceedings that he did.

Dealing now with the second ground of defence that the damage claimed is attributable to fire which is specifically excluded from the policy, and not to an "accident" within the meaning of that word contained in the policy, I agree with my brother Locke that it is unfounded.

The terms of the policy are as follows:—

To pay the Assured far loss on the property of the Assured directly damaged by such accident (or, if the Company so elects, to repair or replace such damaged property), excluding (a) loss from fire (or from the use of water or other means to extinguish fire, (b) loss from an accident caused by fire, (c) loss from delay or interruption of business or manufacturing or process, (d) loss from Jack of power, light, heat, steam or refrigeration, and (e) loss from any indirect result of an accident.

The relevant schedule attached to the policy is the following:—

B. As respects any such unfired vessel, "Object" shall mean the cylinder, tank, chest, heater plate or other vessel so described; or, in the case of a described machine having chests, heater plates, cylinders or rolls mounted on or forming ta part of said machine, shall mean the complete group of such vessels including their interconnecting pipes; and shall also include water columns, gauges and safety valves thereon together with their connecting pipes and fittings; but shall not include any inlet or outlet pipes, nor any valves or fittings on such pipes.

C. As respects any object described in this Schedule, "Accident" shall mean a sudden and accidental tearing asunder of the object or any part thereof caused by pressure of steam, air, gas, water or other liquid therein, or the sudden and accidental crushing inward of the object or any part thereof cause by vacuum therein; and shall also mean a sudden and accidental cracking of any cast iron part of the object, if such cracking permits the leakage of said steam, air, gas, water or other liquid, but leakage at valves, fittings, joints or connections shall not constitute an accident.

If, therefore, the damage claimed is attributable to fire, which is specifically excluded from the policy, the action

[Page 194]

must fail. On the other hand, if the damage is the result of an accident within the above definition, and if it is the direct consequence of such accident, the action must succeed and the appeal allowed.

On August 2, 1942, in the East Room of the oil mill at the appellant's plant in Montreal, some of the employees were in the process of bleaching turpentine in a tank called Tank No. 1. This tank was normally used for bleaching linseed oil. In the course of these bleaching operations, a "sizzling" sound was suddenly heard, coming from Tank No. 1, and which was obviously caused by vapour escaping from the periphery of the manhole door of the tank, and this was followed by the sound of the blowing out of the door under the high pressure of this vapour. The evidence reveals that this vapour in itself was not inflammable, but that it was, when it came in contact with the air. Within a few seconds, a terrible explosion occurred causing to the building extensive damage.

It is the contention of the respondent that the loss suffered by the appellant was not a loss directly caused by accident, there being a nova causa that intervened which was fire, and as the respondent is only liable for direct damage caused by an explosion, it therefore denies all liability. The theory is that, although there has been a minor explosion in Tank No. 1, the vapour that escaped from the tank, coming into contact with the air, was ignited by a fire, which was probably an electric spark, and it was only after the intervention of this new cause that the explosion occurred.

In order to determine this direct cause, it must be kept in mind, as Lord Dunedin said in Leyland Shipping Co. v. Norwich Union Fire Insurance Society 9:—

In other words, you seek for the causa proxima, if it is well' understood that the question of which is proxima is not solved by the mere point of order in time.

In the same case at page 355, Lord Finlay L.C. said that the determining cause of an accident is what "in substance" causes the injury. The damage that may be claimed is the damage which is the "natural consequence" of the accident.

In Cory v. Burr 10, it is said that the proximate cause

[Page 195]

is the direct and immediate cause. In Gordon v. Rimming-ton 11, Lord Ellenborough uses "causa causans" as the equivalent of proximate cause.

It is true that the vapour that escaped from the tank as the result of the explosion, was while floating in the air of the building, suddenly ignited by an electric spark, but I have come to the conclusion that the "causa causans" of the damage suffered, what "in substance" caused the damage, was the explosion in the tank. The last explosion was the natural sequel, the consequence of the original explosion in the tank, which was the main element of causation,

In Leyland Shipping Co. v. Norwich Union Fire Insurance Society, cited supra, a ship belonging to the appellant was torpedoed, while on a voyage from South America to Le Havre. With the help of tugs the ship reached Le Havre, and she was brought inside the outer breakwater, where she remained for two days, taking the ground at each ebb tide, but floating again with the flood. Finally, her bulkheads gave way, and she sank and became a total loss. It was held that the grounding was not a "novus casus intervenions" and that the aggravation of the original injury by the bumping against the quay and the successive groundings, did not convert the partial loss into a total loss. The chain of causation between the injuries caused by the torpedo, and the ultimate sinking of the vessel, was not broken by the series of events which occurred in Le Havre.

In the present case, I have come to the conclusion that there was an unbroken sequence between the explosion in Tank No. 1, which is the casualty, and the ultimate loss. There was not an intervening cause, in which was merged the original casualty.

For these reasons, the appeal should be allowed and the judgment of the trial judge restored with costs throughout.

Rand, J. (dissenting):—I take the circumstances of the loss of the appellant's property to be these. The course of escape of gas generated in the tank by the mixture of turpentine and the other substances, as the pressure mounted, was first by way of the small aperture in the manhole door or the vent at the rear, then between the manhole door, forced outward, and the frame, and finally through the

[Page 196]

manhole when the door was blown off. The sizzling noise was produced in the second stage; and the first explosive sound was the blasting of the door. The gas mixing with the air in the room became combustible and was ignited by a spark probably from an electric mechanism. This burning tended to reach back toward the source of the gas and while its quantity was limited the combustion was relatively slow and presented flames flashing in different directions as it followed the air currents. When the manhole opened the quantity was so great that the rapidity and extent of combustion issued in an explosion. Tongues of flame licked up the thin streams of grayish gas before that point was reached; both gas and flames were seen through both doors by the men working in the adjoining room. There was this fire in the eastern room for a sensible period of' time before the explosion apart from the spark or other source of the original ignition.

The passage of that fire into explosion resulted from the sudden access of the gas; if the slow feed or emission had been maintained or if the peak pressure had been reached before the door gave way, there would have been only the fire. In that case it would ordinarily follow that any damage done by it, either through the burning of property insured or by producing other direct effects, would be fire loss.

Whether the ignition of the gas can be said to have been due to a fire within the meaning of the fire policies ceases, then, to be of importance. There was clearly a secondary stage of fire which superseded the initial cause.

Before deducing the legal consequence from the insurance contracts and the facts stated, I venture to point out the distinction between fire damage and damage caused by fire. An insurance against the former looks to the nature of the loss or destruction; it is damage by burning or combustion only. But insurance against damage by fire treats fire as a cause which in the course of its career may set off other agencies, such as explosion, to bring about damage other than fire to be charged against it. The same consideration arises in exceptions from the main risks; and the question is whether the exception is as to the kind of damage or to the consequences of a certain cause.

[Page 197]

The terms of the policies here are, in this respect, reasonably free from doubt. The parties agree "respecting loss .… from an Accident as herein defined to the object described herein." The Company "agrees to pay for loss ' on the property of the assured directly damaged by such Accident .… excluding (a) loss from fire (or from the use of water or other means to extinguish fire, (b) loss from an accident caused by fire .… , and (e) loss from any indirect result of an Accident." The tank was undoubtedly what is called in schedule 2 an "unfired vessel" and an accident to such a vessel was descirbed as "a sudden and accidental tearing asunder, etc." The bulging of the manhole door and its later blasting, was, in my opinion, a rending asunder within that definition.

The language "property .… directly damaged by such accident" deals with accident as a casual agency and without more would embrace all loss directly resulting from it: loss "from fire" must, I think, be given the same meaning: Stanley v. Western Insurance Company 12; and it is intended to eliminate from the trail of consequences of an accident all those which are to be attributed to the interposition of fire as the efficient factor in a chain of subsequent effects.

This may perhaps be clarified by elaboration. The exception is from a liability for an "accident" and its results. The "fire" must then appear or be involved in those results, otherwise it would be outside the risk assumed; and as the word is used in a causal sense, the exclusion extends to all effects that follow from it as cause: we are to conceive it as a new point of departure, and disregard antecedents. In an ordinary policy against fire, we do not go back for originating causes; what has brought fire about is irrelevant; we take it as if it were a first cause. The same conception is to be given to fire as an exception; when it appears we mark it as a new factor and we are not concerned with what has preceded it. Was it then an actuating agent here? I am bound to say that the answer seems to me to admit of no doubt. It was the flame that set the mixed gases into combustion so great and. rapid as to produce the explosion. Both the gases and the fire were necessary to that reaction, but the fire was the actor in producing it. The problem is

[Page 198]

not one of abstract or philosophical causal determination; we are endeavouring to ascertain the scope of an exception from a risk assumed, the language of which carries the ordinary and popular sense of these phenomena.

Mr. Mann contends that a fire in other than property-insured producing other causes of damage does not entail liability under a fire policy; from which he concludes that, as one or other of the groups of insurance companies must be bound, the disaster must be attributed to the accident. It was, no doubt, a fact that the gas, as substance or property, was not insured against fire; but in the case of Hobbs v. Guardian Insurance Company 13 approved by the Judicial Committee in Curtis's & Harvey Limited v. North British Company 14, neither was the match or the gunpowder; and yet this Court held that the burning powder was fire so as to carry responsibility for the explosion which ensued. I take this decision to mean that fire as a cause of damage insured against is fire in any form which may by its proximate consequence produce loss to the property insured. That is precisely what we have here. But whether liability arises accordingly on the part of the fire insurers is a matter beyond the issues in these proceedings; it is enough that the fire be within the exception of the respondent's contract.

This view differs from that of the Chief Justice at trial in the significance attributed to the flashes of flame previous to the explosion. He considers it too fine a distinction, in relation to the language of the policy, to resolve the developing explosion into stages and to treat the first and second—the ignition and the gas combustion periods—as constituting a "fire" existing as such, to be taken as a cause of new consequence. But that depends on the facts and I am unable to interpret them here as not creating an intermediate state of fire, either of the original gases or in the initial stages of the explosion. Time is significant and explosion was not necessarily involved in the burning gases. The minutes or even seconds which elapsed marked a period not of explosion but of a state of things that, in combination with new elements, led to explosion; the impact of the mass of gas upon the floating fire was the same as the contact of

[Page 199]

the burning match with the powder in Hobbs, supra, and likewise the development of the burning mass into explosion.

The appeal must, therefore, be dismissed with costs.

Estey J.:—The appellant at the trial recovered from the respondent under an accident policy for that portion of loss attributed to an explosion. This judgment was reversed upon appeal 15 and the appellant (plaintiff) further appeals to this Court.

On August 2, ,1942, the appellant in its linseed oil mill in the City of Montreal was filtering turpentine. A No. 1 steam-jacketted bleacher tank (hereinafter referred to as "the tank") was used as part of the apparatus. This tank was located with other equipment, including motors and dynamos, in the east room on the top or third floor of the mill. 850 gallons of discoloured turpentine were poured into this tank, the steam turned into its jacket at a temperature of 145° to 160°F. Then 200 lbs. of "filtrol" and 50 lbs. of "filter cel" were placed in the tank and the agitator therein operated for half to three-quarters of an hour. It is established that this operation of the agitator in the contents of that tank would generate enough heat and pressure to first push the door and permit some vapour to escape through the periphery with a hissing or sizzling noise and then as the pressure was building up rapidly to quickly blow the door open releasing a large quantity of vapour.

In the room adjoining and to the west was other equipment including the filter presses. The men in charge were not satisfied with the turpentine coming through and gathered around the filter presses. In that position they heard a hissing or sizzling noise. One saw "fumes or vapour, then saw fire," another "saw a big flash like fire" and a third was not sure whether he saw flames or fumes in the doorway connecting these east and west rooms. The men all hurried to the fire escape. As they reached the fire escape they heard a "boom" which is accepted as that of the door being blown off the tank. Then as they proceeded down the fire escape they heard an explosion which damaged the roof, walls and windows and which generally

[Page 200]

disturbed the entire structure. The fire followed and the total loss incurred was about $159,724.62. The companies holding the fire insurance have paid that part admittedly caused by fire, but the balance of $45,791.38 is that which resulted from the explosion and which in this appeal the appellant claims under the terms of the accident policy.

The evidence is to the effect that as the vapours escaped through the periphery they were ignited by contact with something in that room, probably an electric switch, motor or dynamo. Whatever it was is described in the proceedings as unidentified, and the fire thus caused was seen by the men as they hurried to the fire escape. One of the experts stated:—

If you have a tank, such as in this case, which is generating vapors quite rapidly and filling alleyways that are 25 or 50 feat long and many feet wide and many feet high full of an inflammable mixture of turpentine vapors and air, it would be a miracle if they did not explode.

It is further explained that these explosions occur in three stages:

In the first stage a flame moves through the explosive mixture at a slow, more or less uniform rate of speed. In the second stage the speed of the flame increases, and the flame may oscillate backwards and forwards in the explosive mixture, and there may be turbulence or a mixing up of the gases in the mixture, and finally there is the third stage in which 'the flame is accelerated in velocity to a great speed and there is usually a loud report and this is the stage termed detonation.

And further:

When an explosive mixture is ignited, a flame forms and moves slowly through the explosive mixture. This slow movement may last for from a fraction of a second to several seconds or minutes, and the rate of velocity usually is from one foot to ten feet per second.

The policy insured the appellant in respect of a loss from an accident to an object. The tank is enumerated among the objects covered by the policy and the blowing off of its door constituted a "sudden and accidental tearing asunder" of the object and therefore an accident within the meaning of the policy.

It is the contention of the appellant that when the large volume of vapor escaped as a consequence of the "tearing asunder" the explosion followed as a direct cause therefrom, while on the other hand, the respondent contends that the explosion was due to the fire.

[Page 201]

Sec. 1 of the policy requires the company:

To pay the assured for loss on the property of the Assured directly damaged by such accident … excluding

(a) loss from fire (or from the use of water or other means to extinguish fire),

(b) loss from an accident caused by fire,

* * *

(e) loss from any indirect result of an accident.

The appellant under the terms of the foregoing sec. 1 in order to recover must adduce evidence establishing that the loss or damage to its property was the direct or proximate cause of the accident. McGillivray, Insurance Law, 2nd Ed., p. 811; Becker, Gray & Co. v. London Assurance Corp. 16. The particular loss we are here concerned with arises out of an explosion the loss or damage from which was not by the terms of the policy specially excluded. The respondent to bring this explosion within the exclusion clause (a) must therefore establish that it was directly or proximately caused by a fire. The issue between the parties is, in these circumstances, what was the direct or proximate cause of this explosion,—the accident or the fire? The position is therefore somewhat similar to that in Leyland Shipping Co. v. Norwich Union Fire Ins. Society 17, where the appellants contended the ship was lost by a "peril of the sea", while the respondents contended the loss was caused by torpedoing for which under the policy they were not liable because of a warranty "from all consequences of hostilities or warlike operations." Lord Dunedin, at p. 363, stated as follows:

But the moment that the two clauses have to be construed together it becomes vital to determine under which expression it falls. The solution will always lie in settling as a question of fact which of the two causes was what I will venture to call (though I shrink from the multiplication of epithets) the dominant cause of the two. In other words, you seek for the causa proxima, if it is well understood that the question of which is proxima is not solved by the mere point of order in time.

In order to have an explosion of the type here in question there must be an inflammable or explosive mixture and it must be ignited. In this case that explosive mixture was the turpentine vapour and the air; it was ignited and in that sense there was a fire.

Everything happened in a very short space of time. The "tearing asunder" of the door released at first a quantity and almost immediately a large volume of turpentine

[Page 202]

vapour into the room. Without the release of the vapour there would have been no explosion. The "tearing asunder" of the door which released such a volume of vapour would appear to have been the direct or proximate cause of the explosion. The presence of the air and ignition were necessary and in that sense causes of the explosion. Seldom, if ever, does an explosion, fire or accident result from one cause. The law, from all the causes leading up to a result, selects that which is direct or proximate and regards all the others as remote. The direct or proximate cause may not be the last, or, indeed, that in any specified place in the list of causes but is the one which has been variously described as the "effective", the "dominant" or "the cause without which" the loss or damage would not have been suffered. In Leyland Shipping Co. v. Norwich Union Fire Ins. Society, supra, the torpedoing of the ship was, though not the last cause, that which was held to be the direct or proximate cause. Lord Atkinson at p. 366 stated:

It is quite true that in the efforts to salve the cargo and the ship her injuries may have been aggravated, but none the less, in my opinion, was the 'loss the direct and immediate consequence of the torpedoing.

In Canada Rice Mills Ltd. v. Union Marine and General Insurance Co. 18, a cargo of rice was damaged by heating. The jury found that the rice was damaged by heating caused by the closing of the cowl ventilators and hatches from time to time during the voyage, and it was held that this was a reasonable precaution, having regard to weather conditions. The policy covered "perils of the sea." The main contest was whether the proximate cause was the "peril of the sea" or the closing of the cowl ventilators and hatches. Lord Wright stated at p. 71:

But it is now established by such authorities as Leyland Shipping Co. v. Norwich Union Fire Society, (1918) A.C. 350, and many others, that causa proxima in insurance law does not necessarily mean the cause last in time, but what is "in substance" the cause, per Lard Finlay (Ibid. 355), or the cause "to be determined by common-sense principles," per Lord Dunedin, (Ibid. 362). The same rule has been reiterated by the House of Lords several times since then, most strikingly, perhaps, in P. Samuel & Co. v. Dumas, (1924) A.C. 431 … Their Lordships agree with this expression of 'opinion, and accordingly are prepared to hold that the damage to the rice, which the jury have found to be due to action necessarily and reasonably taken to prevent the peril of the sea affecting the goods, is a loss due to the peril of the sea and is recoverable as such.

[Page 203]

The foregoing is on the basis that there was a fire within the meaning of the policy. In view of the conclusion arrived at, it is unnecessary to deal with the important question whether the fire was actually a fire within the meaning of the policy or a part of the explosion.

The circumstances surrounding the payment to the appellant by the fire insurance companies of the amount here claimed are such as not to deprive the appellant of an interest sufficient to initiate and carry on these proceedings. Upon this issue I have had the advantage of reading the reasons of my brother Taschereau with which I fully agree.

The appeal should be allowed with costs.

Locke, J.:—This is an appeal from a judgment of the Court of King's Bench for Quebec (Appeal Side) which allowed an appeal by the respondent insurance company from a judgment of Tyndale, J. which had condemned the respondent to pay the sum of $45,791.38 loss occasioned by an explosion on the appellant's premises. Letourneau, C.J., dissented and would have dismissed the appeal.

By the insuring agreement in question, the respondent company agreed with the appellant "respecting loss (excluding loss of the kind described in section II and including loss of the kind described in section IV) from an accident as herein defined to an object described herein occurring during the policy period "inter alia to pay the assured for loss on the property of the assured directly damaged by such accident (or if the Company so elects to repair or replace such damaged property) excluding (a) loss from fire (or from the use of water or other means to extinguish fires), (b) loss from an accident caused by fire, (c) loss from delay or interruption of business or manufacturing or process, (d) loss from lack of power, light, heat, steam or refrigeration, and (e) loss from any indirect result of an accident." By a schedule to the policy the unfired vessels covered were certain objects designated in a further schedule and included a steam jacketed bleacher tank situate in the East room on the third floor of the appellant's factory in Montreal and as respecting any object described in the schedule "accident" was declared to mean:

A sudden and accidental tearing asunder of the object or any part thereof caused by pressure of steam, air, gas, water or other liquid therein or the sudden and accidental crushing inward of the object or any part

[Page 204]

thereof caused by a vacuum therein; and shall also mean a sudden and accidental cracking of any east iron part of the object if such cracking permits the leakage of said steam, air, gas, water or other liquid, but leakage at valves, fittings, joints or connections shall not constitute an accident.

The limit of liability for any such accident to one of the designated objects was $50,000.00.

While it is admitted that there was an accident to the steam jacketed bleacher tank above referred to which was followed by an explosion and by fire, there is disagreement as to just what constituted the accident. Stated briefly the facts are that on August 2nd, 1942, during the currency of the policy the bleacher tank was being used by the appellant company for bleaching a quantity of turpentine for the first time. Theretofore it had been used only for the purpose of bleaching linseed oil and it was attempted to bleach turpentine in the same manner. The process involved placing a quantity of turpentine in the tank together with Fuller's earth and a substance called Filter Cel, heating the mixture mechanically. This work was undertaken apparently without a proper appreciation of the danger involved: the effect of the process was to build up a very heavy pressure within the vessel which first loosened the manhole door of the tank permitting an escape of a quantity of vapour, and then blew off the door permitting the escape of a larger quantity. According to the witnesses who were in the adjoining room on the third floor of the factory, they first heard a hissing or sizzling noise which the learned trial judge considered to have been caused by the vapour escaping from around the periphery of the manhole door which had been loosened by the pressure, and this was followed closely by the sound of the door being blown out by the pressure of the vapour. It was within a matter of seconds thereafter that the explosion occurred, causing the shattering of the upper part of the building in respect of which the appellant's claim is made. While a fire followed which did extensive damage to the appellant's premises the resulting loss, liability for which on the part of the respondent was excluded by the policy, was covered by fire insurance policies and no question arises as to this.

There was conflict in the evidence of the foreman and some of the other workmen who were in the room adjoining that in which the tank was situate as to whether any fire

[Page 205]

was visible before the explosion occurred. The appellant's foreman said that after hearing the sizzling noise caused by the escape of the vapour from the tank, he saw a flash "like a shot of lightning" and immediately shouted to the men to get out, and this was followed by the noise undoubtedly caused by the door being blown off the tank and this promptly by the explosion, the whole sequence of events lasting, according to him, a very few seconds. Others who were present did not see this, but the point is not of importance in view of the fact that the learned trial judge accepted the evidence of Dr. Lipsett and Dr. Lortie, expert witnesses called by the appellant that a flame would undoubtedly be present in the explosive mixture formed by the mingling of the turpentine vapour with the atmosphere before the actual detonation. According to these witnesses, an explosion of this kind, where the mixture is not closely contained within a vessel occurs in three stages: in the first, a flame moves through the explosive mixture at a slow rate of speed, in the second the speed of the flame increases and it may oscillate backward and forward in the explosive mixture and there may be turbulence or a mixing up of the gases, and finally a third stage in which the flame is accelerated in velocity to a great speed and there is usually a loud report, this being termed detonation. The source of the ignition of the mixture however was not shown. Various possible explanations were given by Dr. Lipsett who said that a mixture of turpentine vapours and air such as was present here can be ignited by a source of ignition that is at 584° Fahrenheit and that a piece of iron at that temperature, which would be far below red heat, could ignite it. This witness said that the manhole door might have become heated up beyond that temperature during the chemical reaction in the tank, but that there were many other possibilities, one of the common causes of ignition of inflammable vapours being sparks from electric motors or from switches or machinery or naked lights and that if there is a large volume of inflammable vapour mixed with the air and set loose in a room, it will usually find a source of ignition. As he expressed it where the vapour was released under the circumstances here existing, it would

[Page 206]

have been a miracle if it did not explode. The learned trial judge found that the source of the ignition of the vapour was not proven.

It is the contention of the respondent that the only-accident was the blowing off of the manhole door and that before this had occurred there was a fire burning in the explosive mixture caused by the mingling of the turpentine vapour which had theretofore escaped from the tank with the atmosphere and that accordingly the loss was "from fire" within the meaning of the exception. It was shown by the evidence that after the turpentine, Fuller's earth and Filter Cel had been placed in the tank, steam had been admitted into the jacket surrounding it under pressure to bring the temperature of the mixture up to 165° Fahrenheit. Tests conducted with a similar mixture by Dr. Lipsett disclosed that when these ingredients were heated to this temperature a chemical reaction started which evolved heat, the temperature of the turpentine and the other materials rising at first slowly until a temperature of about 250° Fahrenheit was reached when the reaction became more vigorous and at 315° Fahrenheit the turpentine began to boil, producing the vapours which the witness considered had built up the pressure in the tank which he estimated would have risen to 50 or 60 pounds to the square inch. The manhole door was held closed by a retaining arm which was in turn held in place by bolts passing through lugs on each side of the door. These bolts were shown to have been 3" in diameter and about 9" in length, and tests conducted by Dr. Lipsett showed that with a pressure such as would have been exerted upon the interior of the door such a bolt bent almost 1" and it was the opening caused by the forcing out of the manhole door permitted by the bending of the bolts or one of them that Dr. Lipsett considered to have been the vent through which the first vapours escaped, causing the hissing noise heard by the witnesses. The tank itself was designed to withstand a pressure of 75 pounds, and according to the witness Hazen would withstand about six times that amount but less than this was necessary to force the manhole door partially open. Hazen agreed with Dr. Lipsett that the sides of the door were forced out or lifted by the pressure produced by the vapour and it is apparent that this could occur only

[Page 207]

if the arm or the bolts fastening it were bent or forced outward. The learned trial judge has found that the sequence of events after the first escape of the turpentine vapour was that it became ignited in some unknown manner, a flash or flame being visible in the vapour, the manhole door then blew off and the explosion followed. The definition of "accident" speaks of a "sudden and acci- dental tearing asunder of the object or any part thereof." The word "tearing" is not, I think; one which would commonly be used to describe the shattering of or the distortion of metals which would bend upon the application of sufficient force. It should be interpreted, however, in my opinion, to include a forcing asunder of parts of the object brought about as in the present case by the application of pressure upon the bolt or bolts sufficient to bend them, and forcing the manhole door out of its seating in the wall of the tank, permitting the escape of the vapour. In my opinion, the forcing out of the manhole door and the bending of the bolt or bolts which permitted this and the subsequent blowing off of the door should be treated as the accident and not the latter occurrence alone.

The damage in respect of which the claim is made was not caused by burning. Against this risk the appellant was insured, and the insurance companies have paid the loss. For the appellant it was urged before us that the expression "loss from fire" should be construed as meaning loss from burning only, but I think this contention cannot be sustained and that loss of which fire is the proximate cause is included in the exception. "Loss from fire" in my opinion is not to be construed differently than if the words were "loss caused by fire" and these words have always been construed as relating to the proximate cause. Coxe v. Employers' Liability Assurance Corporation Limited 19, Scrutton, J. The expression "proximate cause" as pointed out by Lord Sumner in Becker, Gray and Company v. London Assurance Corporation 20 is not an ideal way of expressing what is intended: he considered that "direct cause" would be a better expression. In Leyland Shipping Company Limited v. Norwich Union Fire Insurance Society Limited 21, Lord Dunedin, in deciding which of two asserted causes had caused the loss of the vessel, said that

[Page 208]

the solution lay in deciding what was the dominant cause of the two. It was expressed by Lord Wright, in delivering the judgment of the Judicial Committee in Canada Rice Mills Ltd. v. Union Marine and General Insurance Company Limited 22 as what is "in substance" the cause. As pointed out by Lord Shaw in the Leyland Shipping case, (supra), to treat the proximate cause as if it was the cause which is proximate in time is out of the question; the cause which is truly proximate is that which is proximate in efficiency.

The law applicable to the matter appears to me to be accurately stated in Welford's Accident Insurance, 2nd Ed., 178 where the learned author says:—

The operation of the doctrine of proximate cause is not affected by the number of causes that may intervene between the peril and the loss. Thus, a scratch may produce septicaemia which develops into septic pneumonia resulting in death. Nevertheless the death is caused proximately by the scratch. In these cases though the loss is not the immediate result of the operation of the peril upon the subject matter of insurance, there is nevertheless no break in the chain of causation which leads through a succession of causes directly from the peril to the loss. They are so intimately connected the one with the other that but for the operation of the peril the loss would not have happened. The relation of cause and effect is therefore established between them, the intermediate causes are themselves brought into existence by the peril and constitute the instruments by which it produces its ultimate result.

And again at page 184:—

If there is a causal connection between the peril and the loss, the excepted cause being merely a link in the chain of causation inasmuch as it is a reasonable and probable consequence of the peril, the peril is the cause of the loss within the meaning of the policy.

The doctrine of proximate cause is common to all branches of insurance (Welford & Otter-Barry on Fire Insurance, 4th Ed., 259). In the Leyland Shipping case, the steamship Ikaria had been torpedoed by a German submarine off the coast of France: the vessel succeeded in making her way into the port of Havre and was taken alongside the quay in the outer harbour. When a gale sprang up, causing her to bump against the quay the harbour authorities ordered her to a berth inside the outer breakwater where she was moored and remained for two days, taking the ground at each ebb tide but floating again with the flood. Finally her bulkheads gave way and she sank and became

[Page 209]

a total loss. The policy sued upon covered loss by perils of the sea, but contained a warranty against all consequences of hostilities, and the action failed. The repeated grounding of the vessel at ebb tide and the floating again with the flood was admittedly the immediate cause of the bulkheads giving way and the sinking of the vessel, but it was held that the torpedoing of the ship was the proximate cause. Barclay, J., in his reasons for judgment on the appeal in this matter, has said that while the policy insured against the risk of direct damage, the subsequent exclusion of fire would seem to exclude fire even if it was a direct cause of the loss, and considered that the decision in Stanley v. Western Insurance Company 23 applied. But here the loss claimed for is not damage by burning but by the shattering of the premises by explosion. In the Stanley case liability for damage by explosion was excluded, and it was accordingly held that there could be no recovery. Here there is no such exclusion. I agree that loss of which fire is the direct or proximate cause is excluded, but in my view the loss was not so caused.

In the present case it was the application of heat by the introduction of steam under pressure into the jacket surrounding the tank heating the contained mixture and producing the turpentine vapours, the pressure of which first loosened and then blew off the manhole door and it was this accident which was the effective cause of the explosion and the resulting damage. I agree with the learned trial judge that there was no break in the chain of causation which led through a succession of causes directly from the peril insured against to the loss. The flash or flame produced by the ignition of the inflammable vapours was undoubtedly a causa sine qua non, as was the grounding of the vessel in the Leyland case caused by the action of the tide, but this was, in my opinion, one of the two intermediate causes, i.e. the mingling of the turpentine vapour with the atmosphere producing the highly explosive mixture and its ignition from the unknown source brought into existence by the peril insured against and not, therefore, the causa proxima. I find nothing in the decision of this Court in Hobbs v. Guardian Assurance Co. 24, to assist

[Page 210]

the contention of the respondent: there the insurance was against loss or damage by fire and fire was found upon the evidence to have been the proximate cause of the damage.

It was contended in argument before us that the onus was upon the respondent to prove at the trial that the explosive mixture had been ignited by fire and that this had not been done, and further that in any event the flash or flame observed by some of the witnesses prior to the explosion was not a fire within the meaning of that expression as used in the policy but, in view of my conclusion that fire was not the proximate cause of the loss, it appears to me unnecessary to deal with either question.

I have had the advantage of reading the reasons for judgment of my brother Taschereau and I agree with his conclusion that the assignments given by the plaintiff to the various fire insurance companies after the commencement of the action, of which no notice was given to the respondent, do not affect its status to sue.

It was further contended for the respondent that in any event it was liable only for a portion of the loss. This is based upon the fact that the appellant carried at the time of the loss insurance with the Associated Reciprocal Exchanges which covered direct loss or damage by explosion, subject to certain conditions and exclusions, one of these relating to "pressure containers". As to this, I agree with the learned trial judge.

The appeal should be allowed with costs here and in the Court of King's Bench and the judgment at the trial restored.

Appeal allowed with costs.

Solicitors for the appellant: Mann, Lafleur & Brown.

Solicitors for the respondent: Hackett, Mulvena, Hackett & Mitchell.