Gaz Metropolitain Inc. v. Toupin et al.

Gaz Metropolitain Inc. v. Toupin et al. Collection Supreme Court Judgments Date 1973-06-29 Report [1974] SCR 1071 Judges Fauteux, Joseph Honoré Gérald; Abbott, Douglas Charles; Ritchie, Roland Almon; Pigeon, Louis-Philippe; Dickson, Robert George Brian On appeal from Quebec Subjects Torts Decision Content Supreme Court of Canada Gaz Metropolitain Inc. v. Toupin et al., [1974] S.C.R. 1071 Date: 1973-06-29 Gaz Metropolitain Inc. (Defendant) Appellant; and Roland Toupin (Plaintiff); and Jean Chabot (Defendant) Respondents. 1973: May 14, 15; 1973: June 29. Present: Fauteux C.J. and Abbott, Ritchie, Pigeon and Dickson JJ. ON APPEAL FROM THE COURT OF QUEEN’S BENCH, APPEAL SIDE, PROVINCE OF QUEBEC Negligence—Installation of gas appliances—Explosion—Damage—Subsidization program—Contract between plumber and gas user—Liability of gas company. Respondent Toupin had work done on his house, and gave an order for work to respondent Chabot, who was to instal, inter alia, a gas water heater and a gas space heater. Under a subsidization program, the appellant company undertook to subsidize the costs of installation up to a fixed amount stated in a form describing the gas appliances ordered and listing the installation requirements. In the kitchen where these appliances were to be installed was an old gas meter; it was put inside a cupboard and attached with the necessary connections, to a “meter bracket” furnished by appellant and equipped with a supply pipe at one end and a feeder pipe for the appliances at the other. The meter was 37 years old and had been inspected 6 years earlier. Moreover, a municipal inspector checked the distribution pipes and indicated he was satisfied that there was no leakage. The joints of the bracket and meter were also checked by respondent Chabot after he completed his work. In the hours that followed no gas was seen escaping until the next day, when an odour of gas accompanied by a hissing noise attracted the attention of one of the tenants. Appellant had barely had time to notice a leak on the meter when a violent explosion occurred, causing serious injuries and considerable damage. Appellant and respondent Chabot were jointly and severally condemned to pay damages to Toupin, and this judgment was upheld on appeal. Hence the appeal to this Court. Held: The appeal should be dismissed. [Page 1072] Per Fauteux C.J. and Abbott, Ritchie and Dickson JJ.: The appeal should be dismissed for the reasons indicated by the learned trial judge. Per Pigeon J.: The first question is whether the company had assumed a responsibility by contracting in the manner described. There is no indication that it undertook any other obligation than to pay the cost of the work. The contract was made between Chabot and Toupin. None of the necessary elements to conclude that Chabot was appellant’s employee or servant were found in the evidence. A contract for services was involved here. So far as the duty of supervision is concerned, the method provided by legislation for dealing with this danger is a compulsory inspection by municipal or provincial officers. This inspection was carried out by the proper officer, and it cannot be said that a public utility has an obligation to provide an inspection service itself in addition to the government service. In order for liability to attach there must be found a causal link between the fault and the damage. As to the origin of the explosion, the trial judge correctly found that the gas leak was traced to the inside of the kitchen cupboard, and he was also correct in deducing that “the said leakage was from a component part of the gas circulation system inside the cupboard”. However, his deduction that the leakage of gas probably did not originate in the meter, which had been checked for leaks in appellant’s shops a little over six years before the accident, does not seem entirely logical. If the company was obliged to check not only the accuracy, but also the absence of leaks, in all gas meters at least every six years, it was because there was a fair likelihood that unsuspected cracks or leaks might occur. The meter should not have been in operation without another check for leaks. On the facts which he accepted as proven before him, the trial judge ought to have found that the gas leak from which the damage suffered by plaintiff originated was due to a defect in the meter. As this meter was appellant’s thing it must necessarily be held liable. The crack in the meter was due to the thing itself, and the gas leak and explosion were its direct and foreseeable consequences. Appellant must accordingly bear full responsibility towards respondent. [Page 1073] With regard to respondent Chabot, there is no need to consider whether he did commit a fault which caused the damage, as he has discontinued his appeal. APPEAL from a judgment of the Court of Queen’s Bench, Appeal Side, province of Quebec[1], affirming a judgment of the Superior Court. Appeal dismissed with costs. G. Emery, Q.C., and G. Doss, for the defendant, appellant. P. de Grandpré, Q.C., for the plaintiffs, respondents. The judgment of Fauteux C.J. and Abbott, Ritchie and Dickson JJ. was delivered by ABBOTT J.—The facts and judgments in the Courts below are fully summarized in the reasons of my brother Pigeon and it is not necessary for me to repeat them. The appeal turns on issues of fact and since I am in agreement with the reasons and conclusions of the learned trial judge which were unanimously affirmed by the Court of Appeal, I would dismiss the appeal upon that basis. PIGEON J.—This appeal is against a decision of the Court of Appeal of Quebec1 affirming a judgment of the Superior Court condemning appellant company jointly and severally with respondent Jean Chabot to pay respondent Roland Toupin damages amounting to $59,429.44 and apportioned equally as between the two defendants. Jean Chabot desisted from his own appeal and was not represented at the hearing. Toupin was the owner of a pretty old three-storey house with a flat on each floor. He occupied the ground floor and had a tenant, Gaston Richard, on the top floor. The other flat, the second storey as it is called in the evidence, was vacant, and Toupin had decided to turn it over to his daughter, who was soon going to be married. In the interim he was having certain work done on it. He had gone to respondent Chabot, who is a plumber, and had ordered, in [Page 1074] addition to a sink, a gas water heater and a gas space heater. To promote increased use of gas appliances appellant company had what it called a subsidization program. Under its corporate name at that time it had distributed to some plumbers, including Chabot, forms headed: [TRANSLATION] QUEBEC NATURAL GAS CORPORATION SUBSIDIZATION—JOB INFORMATION SHEET These contained blanks for the customer’s name, a description of the heating equipment sold, and finally the “installation requirements”. Toupin signed one of these forms, describing the two gas appliances ordered and listing “installation requirements” amounting to $177.48. Above his signature is the following statement: [TRANSLATION] IT IS UNDERSTOOD THAT THE QUEBEC NATURAL GAS CORPORATION WILL SUBSIDIZE $……. ONLY. BALANCE IS TO BE OBTAINED FROM PURCHASER BY INSTALLER AFTER COMPLETION OF INSTALLATION. This form was given to a representative of appellant, one A. Beaudet, who signed alongside the words “Counsellor’s Signature”. It is not denied that this implied an undertaking to pay all the installation expenses, and that Chabot was told of this. In a corner of the kitchen of the vacant apartment was an old gas meter dating from 1926, connected by piping to a main gas pipe located alongside the chimney built into the lateral party wall of the house. Toupin wanted to place the stove in this corner in front of the chimney, and the meter inside a cupboard he planned to build along the balance of the wall. Chabot, or rather a workman employed by him, therefore began by removing the meter and the piping as far as the main pipe, in order to replace it with new piping four to five feet in length, at the end of which he put, with the necessary connections, what was described in the evidence as a “meter [Page 1075] bracket”. This “bracket” is a fixture placed horizontally, with a stop cock at one end connected to the supply pipe, and a connection at the other end to the feeder pipe for the appliances. On either side below this bracket was a coupling for the meter inlet and meter discharge. The old meter was not installed in this way: it was coupled with lead tubing, which was no longer permitted. The new bracket, previously used elsewhere, was the property of appellant, who had left a supply of such brackets with Chabot. The latter completed installation of the gas piping designed to serve the appliances, and took care to plug the meter couplings under the bracket. Later on came an employee whose job it was to read meters; he sealed the stop cock so that it could not be opened without breaking the seal, read the meter and made a report to the company. Subsequently a municipal inspector checked the distribution pipes using compressed air, and indicated he was satisfied that there was no leakage. After this Toupin himself connected the stove. A few days later, about 2 P.M. on June 4, 1963, Chabot connected the water heater and the space heater, and installed the meter putting into the couplings washers provided by appellant company. At that time, the bracket was enclosed in a cupboard, and the two gas pipes entered it through an opening in the side, about three inches by eight inches large. In order to instal the meter Chabot had to enlarge this opening at the top somewhat. He broke the seal and opened the stop cock with a key he had made. Using soapy water, he made certain the joints of the bracket and meter were not leaking, lit the pilot lights on the stove and the water heater, and got both appliances to operate. He then worked on the sink and left the premises at about four-thirty. [Page 1076] During the course of the evening respondent Toupin’s daughter went to her future flat three times, and no one observed gas escaping. The next day, shortly after noon, the tenant above smelled an odour of gas while at lunch. He knocked at Toupin’s door, and accompanied the latter to the vacant flat. There was a strong odour of gas and a hissing noise could be heard. The two men went to the cupboard in the kitchen after checking the water heater. Toupin had barely had time to notice a leak on the meter when a violent explosion occurred, causing serious injuries and destroying the house along with the adjoining property. At trial each party called an expert witness. The trial judge rejected in toto the opinion expressed by the company’s expert, and rightly so. Plaintiff’s expert, a chemist named Bernard Péclet, testified essentially as to the detection of a gas odour and the conditions under which ignition and explosion of gas mixed with air occur. W.C. Viner, a mechanical engineer, testified for defendant Chabot. He expressed the opinion that the probable cause of the accident was a leak from the old meter. On this crucial point of the case the trial judge said: [TRANSLATION] 5. Having said this, it seems logical and reasonable to formulate the following four propositions: (a) There was no leakage of gas which could be smelled before 11 P.M. on June 4. Hence the presumption that everything was in order. (b) No one knows when this leakage of gas began. By midday on June 5, however, it had reached a point where Richard, on the third floor, noticed it. (c) The leak was traced to the inside of the cupboard in the kitchen. That is the version given by plaintiff, and I was greatly impressed with his sincerity, candour and good faith. I have no hesitation in accepting this version. I would add that in addition it is substantially identical to that of Richard (see the examination on discovery of the latter [Page 1077] in Richard v. Toupin et al. (M.S.C. 643 512) which was filed in the record by consent). Hence the presumption that the said leakage was from a component part of the gas circulation system inside the cupboard. (d) These components consisted of part of the gas supply pipe, the meter bracket, the meter inlet coupling, the meter itself, the meter discharge coupling and part of the feed or distribution piping. Hence the presumption that there was a failure in one or more of these components at a given, but unspecified, moment on the night of June 4-5, or on the morning of June 5. 6. At this point, faced with the brutal and unquestioned fact of the explosion (“res ipsa loquitur”), I pass into the realm of probability. In his factum plaintiff mainly suggests two probabilities. He says (pp. 8 to 11): “…On the whole there is no doubt, in view of the evidence, that the weakest part compared with the other components is the meter, not as the witness Viner pointed out from the standpoint of pressure, but from the standpoint of possible cracks developing as a result of the greater fragility of its joints, which are necessarily subject to shocks and vibrations even during normal use of the appliance. And this, we submit, is all the more true as this meter had been in operation for 37 years, and had not been inspected since January 24, 1957, contrary to the provisions of the act which requires the Corporation to submit its gas meters to the Department of Trade and Commerce every six years so they can be checked for calibration and leakage (Mr. Dumas). …the witness Laframboise told us of having been suddenly confronted with a leak on a gas meter in a restaurant on Ontario Street in December 1963. And in the opinion of the witness Viner it would be the meter here also. However, plaintiff respectfully submits that the witness Viner was perhaps somewhat hasty in dismissing the possibility that the leak might also have come from the couplings to the meter, particularly in view of the testimony of Mr. Chabot on this point. It is true that the latter is Mr. Viner’s client. In fact, at one point in his testimony Mr. Chabot indicated what a delicate matter it was to connect up a meter; because [Page 1078] the couplings are fragile, they break easily if tightened too hard with a wrench, and he happens to have broken some on several occasions. We respectfully submit that in the circumstances a break in a coupling is just as likely as a leak in the sides of the meter, and this is why Mr. Péclet goes that one step further than Mr. Viner in his testimony. Taking into consideration all the facts proved we feel that the meter as well as its couplings must be regarded as the most probable locations of the leak. I accept without hesitation the second of these probabilities. I feel it is reasonable and supported by the evidence. With regard to the second (let us read “first”), I would say that it appears to me to be much more in the realm of possibility than of probability. The dividing line between these two is obviously very fine; but I cannot be satisfied that it is probable that the gas leak was due to a crack which suddenly developed or appeared in the meter itself. I have no reason to discount the testimony of the shop superintendent of defendant Corporation, Mr. André Lodomez, who explained (Lodomez, April 25, 1968, pp. 65 and 66) what was involved in the “inspection” of a meter, and I quote: “…At that stage when the meter was brought to the shop, the first thing that happened was, it (the meter) was tested to see how accurate it was, and then it was taken completely apart, the mechanism was repaired, it was put together again and the casing securely closed, and it was subjected to a pressure of three and a half (31/2) pounds for one minute, under water…” I think I can assume that the meter, though it was in more or less continuous operation for thirty-seven years, was in perfect working order after its last “inspection”, that is on January 24, 1957. Now, there is not a scintilla of evidence before the Court that in the six years and four months (more or less) for which it was used after January 24, 1957, and up to the day in early May 1963 when it was removed, and placed in a cupboard by plaintiff, the meter gave any indication of weakness, faultiness or wear. A priori, therefore, I do not consider it probable that this meter could have “developed weak spots” so [Page 1079] suddenly and unexpectedly, when on the day preceding the explosion, once it had been installed by defendant Chabot, it successfully underwent the soap test. In contrast with the other components, the meter was not as such subjected to any particular work by hand. The only time there was human intervention in its case was when Chabot placed it in the cupboard and connected it to the bracket. There was no suggestion that he dropped it or at some point knocked it on one of the corners, or anywhere else. In such circumstances I must eliminate the meter in my search for the probable cause of the explosion, and I attribute the latter entirely to the fault, negligence or imprudence of defendant Chabot. The trial judge then went on to say that the company “cannot avoid liability under the pretext that every ‘installer’ is an independent contractor over whom it has no control”, and taking into consideration the company’s familiarity with the dangers of the gas which it sells, he concluded as follows: [TRANSLATION] I conclude, in the specific instance before the Court, that defendant corporation must answer for the consequences of Chabot’s fault or negligence, jointly and severally with the latter. Finally, having assessed the damages at $59,429.44, he set the share of each defendant at half this sum. The Court of Appeal upheld the decision and endorsed, in effect, the reasons of the trial judge. In this Court appellant essentially raised two points. First, that it is not responsible for the fault of Chabot, as he is an independent contractor and not a servant. Secondly, that, if it is found to be responsible for the fault of Chabot, there is no legal basis for the division of liability with the latter, and it should recover the whole amount of the condemnation, not just half. It was argued on behalf of Toupin that if the company was not responsible for Chabot’s fault, it should still be held liable because the accident [Page 1080] was caused by a crack in the old meter, appellant’s thing, for which it is responsible as owner. Appellant’s two points appear to me well founded. As it says in its factum, citing inter alia the decision of this Court in Quebec Asbestos Corp. v. Couture[2], not one of the factors essential to a contract of lease and hire of personal services is found in the case at bar, while all the necessary elements of work by contract for services (contrat d’entreprise) are there. Chabot’s work was for a fixed price, without regard to time or materials used. He had his own employee on the premises. He supplied all the equipment as well as his own tools. He was master of the methods used, he could work when he pleased provided he finished within a reasonable time. In truth, neither the trial judge nor the Court of Appeal found in the evidence the necessary elements to conclude that Chabot was appellant’s employee or servant. The trial judge took the following approach: [TRANSLATION] In my view, the defendant corporation—a well-known public utility—who during an expansion campaign made an offer to the public to defray, in whole or in part, the cost of installing the necessary piping for the use of its product, who distributed forms for this purpose to selected “installers”, who relied on the latter regarding the “installation requirements” (Exhibits P-16 and P-16A), and who reserved the right to approve or disapprove the information submitted by the said “installers”, and so to permit, or not permit, the work to be done on a subsidized basis (provided it was done by the “installers”), cannot evade liability on the pretext that each “installer” is an independent contractor over which it has no control. I would go further. During Péclet’s testimony (April 24, 1968, pp. 15 and 16), defendant corporation admitted… “that it knew the serious dangers likely to result from a leak of natural gas from the pipes in which it was contained”. [Page 1081] Does this admission entail obligations based on the most elementary rules of caution? I readily concede that defendant corporation probably was unable to supervise itself and through its own employees the installation of pipes in the homes of each of its customers. At the same time, in view of “the serious dangers likely to result from a leak of natural gas”, was it not obliged, in offering to subsidize these installations, to see to it that the “installers” followed approved practice and took all the required precautions? With respect, I must say that this reasoning appears to be erroneous. The question is not whether the company could “evade” liability, but whether it had “assumed” such liability by contracting in the manner described. At the hearing counsel for the respondent maintained that appellant had bound itself to provide the installation. This is not what the evidence and supporting documents show. There is no indication that it undertook any other obligation than to pay the cost of the work. This was ordered by respondent without any intervention by appellant. The contract was made between Chabot and himself. All that appellant did was to intervene by agreeing to pay the cost. It was argued that appellant could choose the contractor, but this is not established by the evidence. It is quite true that Chabot stated that at one point an employee of appellant had told him that the work should be done by someone else, but finally the employee had admitted that he was in error in saying this. Turning now to the duty of supervision, I fail to see any basis for this. It is quite true that gas is a dangerous substance, but the same applies to electricity and many other things. The method provided by legislation for dealing with this danger is a compulsory inspection by municipal or provincial officers. Here the evidence is that the inspection was performed by the proper officer. I cannot agree that, under such circumstances, a public utility has an obligation to provide an inspection service itself in addition to the government service. Moreover, there is no indication that the municipal inspec- [Page 1082] tor did a poor job. Nothing suggests that any other inspector employed by appellant would have done any better. In order for liability to attach there must be found a casual link between the fault and the damage. Here there is nothing in the evidence to show that another inspection would have prevented the damage. As we have seen, the agreement concluded consisted of a contract for services between Chabot and respondent. The latter was the owner of the building in which the work was to be done, the work was ordered by him and he was thus in full control. Appellant only undertook to pay the cost, and so did not place itself in the position of one who has work performed. For this reason there could be no question of the kind of liability dealt with in the decision of the Court of Appeal in St. Louis v. Goulet[3]. The owner of some land had given a contractor the job of clearing it and burning the felled brush. He was held responsible for the damage caused to a neighbouring property when the flames spread. A similar decision was given in City of Sherbrooke v. Roy Limitée[4], which concerned the death of a workman caused by dynamiting work carried out by a sub-contractor. In both cases the general contractor was held liable. That is not appellant’s position here. This conclusion, like that of the absence of any master-servant relationship, is strictly a question of law: it derives not from the assessment of the evidence but from the legal consequences that must be drawn therefrom. It is therefore necessary to consider the alternative argument submitted by respondent requiring a consideration of the inferences drawn by the trial judge as to the origin of the explosion, which in no way implies questioning the facts on which he based those inferences. In stating the facts I have considered it desirable to quote, without deleting anything but a portion of a lengthy citation, the entire reasons of the trial judge on the question of liability. In his fifth [Page 1083] paragraph he summarized the facts proven, and I am in full agreement with him. He correctly found that the gas leak was traced to the inside of the kitchen cupboard, and said “That is the version given by plaintiff, and I was greatly impressed with his sincerity, candour and good faith”. He was also correct in deducing therefrom that “the said leakage was from a component part of the gas circulation system inside the cupboard”. However, he was only making a deduction when in paragraph 6 he found that the leakage of gas did not probably originate in the meter; and the only reason he gave for reaching this conclusion is the fact that the meter had been checked for leaks in appellant’s shops a little over six years before the accident, namely on January 24, 1957. This deduction does not seem to me entirely logical. It is quite true that there was no evidence that the meter had given any indication of being worn or defective. But why is the company obliged to check not only the accuracy, but also the absence of leaks, in all gas meters at least every six years, if not that there is a fair likelihood that unsuspected cracks or leaks might occur? Here the meter concerned should not have been in operation without another check for leaks. Moreover, the witness on whom the trial judge based his finding on the location of the leak in the cupboard, the plaintiff Toupin, placed this leak not only in the cupboard but in the meter itself. He testified as follows in cross-examination on his examination on discovery. [TRANSLATION] Q. Why did you go to the meter? A. Because I head a hissing noise from there. It was easy to see that it was coming from there. It was quite pronounced. … Q. Did you feel, on the skin of your hand, not by smelling, the place the jet of air or gas was coming from? A. No, I jumped immediately. The key, I tried closing it off, when I saw this was impossible I [Page 1084] moved back, as I was moving to the door I held the door and that exploded and, then, it was a jet of flame that came out from the meter, that was rushing out like a flame thrower, that burned me all over, roasted me all over, and then I lost consciousness, I was thrown against the wall. … Q. Can you say from what part of the meter the noise came? A. Well, I was looking at the meter, I was hearing the noise coming from the meter. From what part exactly, it is hard to say. Q. But it came from the meter? A. Yes. Q. It wasn’t coming from the inlet or discharge connection? A. No, sir. Q. It was coming from the casing itself? A. Yes, from the casing itself. Counsel for the appellant argued that the trial judge did not omit to consider this important piece of evidence, since he did cite, in his statement of the facts, the second of the answers above quoted. In my view this is not conclusive since, if the trial judge had in fact stopped to consider plaintiff’s testimony from this angle he would necessarily have had to take into consideration the answers to subsequent questions in which plaintiff was asked to describe clearly the location of the leak by distinguishing between the meter and its connections. Further, according to the rule laid down in Dorval v. Bouvier[5], for a conclusion of fact to be given its full effect it must be adequately reasoned. Here, in order to conclude that this evidence was not wrongly disregarded, we need to be able to see why he did not take it into account. We cannot assume that the trial judge refused to believe this testimony, since on the contrary he regarded it as wholly sincere. Nor can it be because the judge regarded such an occurrence as implausible, since he indicated clearly that such a leak is perfectly possible in a meter. This is why he referred to the dividing line between possibility and probability, and went on to say, [Page 1085] “I cannot be satisfied that it is probable that the gas leak was due to a crack which suddenly developed in the meter.” With respect, I cannot see how the trial judge came to this conclusion. Assuming for a moment that he did not wish to accept the testimony of defendant Chabot and of his employee Laframboise, who said they had witnessed such leaks, I see no reason for not taking into account the deposition of the expert witness Viner, who was not contradicted on this point by anyone, and who explained how a leak may suddenly occur in one of these meters. Q. Would you draw a relationship between the age and the time, the duration of the time the meter has been used and the likelihood of its springing a leak? A. Well, I think the most potent source for leakage from this type of a device, a sheet metal device whose sides are soldered together, is probably the random bangs and knocks and stresses for installation for normal use from the gradual crystallization of the soldered material over the passage of time. From all the stresses to which it is subjected. Ideally, it is in a fairly sheltered position, but over the long run obviously some of these defects occur and hence the gas company sets up a regular maintenance program which, as we have heard, removes gas meters every six years for inspection. I wouldn’t want to speculate whether six years is a suitable figure, or whether sixteen years is, but I think when the gas company picks on a six years figure, it is probably based on sound statistical evidence from their records that, if they pick a six years period, they have a fighting likelihood of catching most of the problems which they know will occur sooner or later. Furthermore, before concluding that the leak was more likely to have occurred in the remainder of the installation inside the cupboard, rather than in the meter itself, should not some consideration have been given to how the other part of the installation, apparently in good order [Page 1086] until eleven the previous evening, could have suddenly become defective during the next twelve or thirteen hours? Why, I wonder, in making a choice between two possibilities, was the same question not posed for each of them? Nowhere in the evidence is there any hypothesis to explain how a defect in the installation under the cupboard, other than in the meter, could have resulted in what was seen to happen, namely, an absence of any apparent leak or odour until eleven the previous evening, but at about noon the next day a huge leak accompanied by a pronounced odour and a hissing noise audible throughout the apartment. It was said that the couplings break easily if tightened too hard, but there was no suggestion as to how they could have broken twelve or twenty hours later. The witness Cockburn was the only one who put forward a theory to explain these facts. This was that the defect in the installation occurred at the junction of the branch piping with the main gas pipe. He suggested that this connection cracked when Chabot lifted the pipes slightly where they entered the cupboard, in order to be able to place the meter under the bracket. This theory was correctly rejected by the trial judge, who concluded that the leak had to be located in the cupboard, not at the junction of the branch piping with the main gas pipe. Moreover, this theory required that the main gas pipe be inside a partition, and the contrary was established. The fact remains, therefore, that no one suggested any hypothesis to explain how the installation could suddenly have become defective inside the cupboard, after 11 P.M. on the evening of June 4, unless by a sudden crack in the weakest part, the meter. In other words, a defect in the meter was the only probability established, and it was also the only one established by the direct evidence of what was seen by plaintiff. For these reasons, therefore, I must conclude that the trial judge erred in the deductions he made from the evidence, and that, on the facts which he accepted as proven before him, he [Page 1087] ought to have found that the gas leak from which the damage suffered by plaintiff originated was due to a defect in the meter. As this meter was appellant’s thing it must necessarily be held liable. The crack in the meter was due to the thing itself, and the gas leak and explosion were its direct and foreseeable consequences. Appellent did not even attempt to show that it was unable to prevent the ensuing damage, which in the circumstances is not surprising. As respondent Chabot discontinued his appeal there is no need to consider whether he, also, committed a fault which caused the damage. The conclusion that the leak must be attributed to a crack in the meter in no way implies that he did not. However, it does justify imputing to the appellant full responsibility towards respondent. For the foregoing reasons I conclude that the appeal should be dismissed with costs. Appeal dismissed with costs. Solicitors for the defendant, appellant: Blain, Piché, Godbout, Emery, Blain & Vallerand, Montreal. Solicitors for the plaintiff-respondent, Roland Toupin: Deschenes, de Grandpré, Colas, Godin & Lapointe, Montreal. Solicitors for the defendant-respondent, Jean Chabot: Provost, Favreau & Godin, Montreal. [1] [1971] Que. A.C. 393. [2] [1929] S.C.R. 166. [3] [1954] Que. Q.B. 185. [4] [1966] Que. Q.B. 239. [5] [1963] S.C.R. 288.