Reliance Insurance v. Pyrofax Gas Corp.

BATTISTI, District Judge. On the morning of March 21, 1961, a fire occurred at the Marineland boat marina located on Harsens Island, Michigan. A large steel building used for the storage of boats together with an attached small frame office building were completely destroyed. The fire was discovered at approximately 6:00 a. m., and, shortly thereafter, when the fire department arrived, the frame office building had already burned to the ground and the main steel building was a mass of flames. The large metal building contained no heating facilities. However, in the frame office there was a Coleman Wall Heater attached to an outside wall between 2" X 4" studs. This wall was of lapjoint wood construction with asbestos shingles on the outside and a sheet of asbestos (thickness y%") on the inside immediately behind the heating unit. The fuel source for the Coleman Wall Heater was a low pressure propane gas system known as the Pyrofax RX System. It was installed for the Pyrofax Gas Corporation by Jerry Ritter Appliances, Inc., on December 2, 1958. — 604 — On October 6, 1961, the owners of Marineland and various insurance companies as subrogees of their insureds filed the consolidated actions which are the subject of this appeal. In these actions the plaintiffs alleged that the fire at Marineland was caused by the defendants’ negligence and breach of implied warranty in connection with the design, manufacture, installation, and repair of the Pyrofax RX System.1 On December 18, 1964, a jury returned a general verdict in favor of the plaintiffs, and judgments were entered against Pyrofax Gas Corporation and Jerry Ritter Appliances, Inc., on December 22, 1964. Defendants’ motions for judgment notwithstanding the verdict and for a new trial were denied on May 27, 1965. On June 21, 1965, the District Court denied plaintiffs’ motions for interest from the date of the loss to the date of the entry of judgment. The defendants appeal from the judgment entered on December 22, 1964, and the order denying the motions for a new trial and for judgment notwithstanding the verdict. The plaintiffs appeal from the order denying interest from the date of the loss to the date of the entry of judgment. In this appeal the defendants raise essentially three points. They are: (1) That the plaintiffs’ crucial opinion testimony was based upon facts or inferences which, by uncontroverted reasonable evidence, were shown not to exist; and, therefore, the District Judge erred in denying their motions for a directed verdict and for judgment notwithstanding the verdict. (2) That the District Judge committed prejudicial error in allowing the jury to consider whether there was compliance with the Michigan Liquid Petroleum Gases Regulations. (3) That the District Judge committed prejudicial error in refusing to give certain of the defendants’ requested instructions and, further, in giving certain of the plaintiffs’ requested instructions. As previously noted, plaintiffs urge that the District Judge erred in denying interest from the date of the loss to the date of the entry of judgment.2 The Pyrofax RX System is designed to supply low pressure gas to an appliance such as a Coleman Heater. It consists of cylinders which are connected by a manifold leading to a low pressure regulator. The cylinders are partially filled with liquid propane. Through molecular action caused by the difference in the temperature of the liquid propane and the air in the cylinder, the liquid assumes a gaseous state and rises to the manifold where it is available for use as a high pressure gas. The regulator functions not only to allow gas to pass to the appliance, but also to reduce the high pressure gas (1300 water column inches) in the manifold to a low pressure gas (11 water column inches). The basic theory underlying plaintiffs’ case is that the fire at Marineland originated as a result of the Pyrofax RX System allowing an excessive amount of fuel to reach the Coleman Wall Heater. In this regard, there is no dispute that the defendants knew that a serious fire hazard would be created if substantially more than 11 water column inches of pressure were allowed to reach the heating unit. Thus, in designing, manufacturing, installing, and maintaining RX Systems, it is clear that the defendants undertook to provide parties such as Ma-rineland with a propane gas system which would allow only slight deviations from the requisite 11 water column inches of pressure. — 605 — Plaintiff’s expert witness, Professor Edwin Harold Young, testified as follows with regard to the effect of an excessive amount of fuel reaching the heating unit: “A. The effect — well the effect would be that it would take in, through the burners, considerably more vapor than was intended. Q. (By Mr. Johnson): And what would be the effect in the appliance; what could be the effect? A. It would over heat. The Court: You mean there would be more fire ? (598) The Witness: There would be more fire; hotter fire; more fire. Q. (By Mr. Johnson): Could there be sufficient fire, Doctor Young, to melt the appliance itself? A. Certainly not the cast iron or the steel components. Propane fire you could get a hot enough flame to melt brass fittings.” (R. 384 b) H. Emerson Thomas, one of the defendants’ expert witnesses, testified to the same effect: “Q. And if propane gas comes into that burner in excess of 13 inches of water — let’s say, it comes in at — let’s take ten pounds. That would, roughly, be about 275 inches of water; is that right? A. That’s right. Q. If it comes in at that pressure, it is going to cause that flame to shoot up? (1471) A. There would be a larger flame, yes. Q. And if it ever went in at, say 55 pounds or slightly over 1500 inches of water, would it not engulf the entire burner and the entire bottom section of the furnace? A. Yes, it would; and it would then melt out the control equipment and possibly the orifice in the unit because you would be getting an excessive heat throughout the whole unit.” (R. 832 b). Professor Young testified that after the fire the physical condition of the Coleman Heater indicated that it had been subjected to the high temperatures of a propane fire. For example, he stated: “Q- (By Mr. Johnson): Dr. Young, have you had occasion to examine Exhibits 58 and 59? A. Yes, I have. Q. And * * * did you find and can you show us evidence on those two exhibits or either of them of a propane fire? A. Yes, I can. I would like to point out to the jury that the furnace is upside down as shown here, that this is the floor edge of the furnace, * * * ****** The Witness: * * * As a metallurgical engineer, several things are quite apparent. The first is that it’s quite obvious that the parts have been in a fire of some high temperature or magnitude. (668) The oxide scale — There are three oxides of iron. FEO, FE 203 and FE 304. The Court: Is that the formula for propane gas? Mr. Vandeveer: I understood they are oxides. The Witness: They are oxides. FEO, FE 203, and FE 304. They are yellow and a reddish brown and also almost red-black practically. It depends upon the exact temperature of which they are formed. ****** The Witness: * * * The lower portion of these two exhibits have obviously been heated to a much higher temperature than the rest. By that I mean the parts in which we see metal completely oxidized off. Now the thing that is interesting about that is the fact that the temperature levels to which you would have to heat iron of this type — it’s a mild steel — for this to occur would be at temperatures considerably higher than — 606 — that which you would obtain from wood. (669) You get upper limits for burning wood 1200 to 1400 degrees fahren-heit'. This is (sic) obviously been heated well over 1600 degrees fahrenheit. ****** Now, if we have a hydrocarbon that is burning, there’s a compound of oxygen and hydrogen, such as propane, burning with air. It will reach a much higher flame temperature, and that flame temperature is determined by the air-fill ratio; that is, the amount of air that is in contact with the amount of fuel burning, the availability (670) of the air to it; so an examination of these parts would indicate that a wood, fire could not have burned away — oxidized away. * * * , * * * The condition of the parts indicate to me a hydrocarbon fuel-fed fire of such a nature that temperatures could be reached such that this degree of damage to the iron can occur.” (R. 424-426) Professor Young also testified that a brass orifice which evidently melted during the fire could not have melted under the temperatures which would have been generated by a wood burning fire. (R. 429 b) Relative to Professor Young’s opinion that the condition of the heating unit indicated that it had been subjected to a propane fire, Mr. H. Emerson Thomas stated: “Q. Now, there was comment testimony in this case that some of the metal on the heater which consists of— I take it, the Exhibit 59 — evidence propane fire, as well as Plaintiffs’ Exhibit 58; did you examine these pieces of metal? A. Yes, I have. Q. Do you have an opinion with regard to that subject matter? A. Yes, I do. ****** Q. And what is that opinion, sir? A. I have in investigating and in my consulting business, many, many fires found many heaters, as well as other types of things that are made of sheet metal, and have seen the same typical (1392) results on the sheet metal, which has been painted or not painted. As I have observed on these two Exhibits, and in many of those inspections that I have made, the evidence was very definite that such damage was caused by a wood or other fires, and not a propane fire, and it is hard to tell the difference between them.” (783-784 b) Another of the defendants’ experts, Mr. Edward R. Biggins, offered a similar view: “Q. Sir, as a metallurgist, can you give us an opinion where by examination with the naked eye you can determine the exact substance that caused certain evidence of burning? In other words, if you examined an object which has been in a fire, can you, with your naked eye, tell what the source of the fire was, as between gasoline, and propane, and paint, and other items that may burn at a high temperature ? A. I don’t believe it could be done. Q. Can you only tell whether it was a high temperature or not? A. You could tell if it was subjected to a fire, but as to the (1861) exact temperature or the approximate temperature would be very difficult to do. Q. Is it possible with a naked eye to say exactly what the substance was that burned, and burned the object being examined? A. I would say it was impossible.” (R. 1073 b) While Mr. Thomas initially testified that propane would not have caused the brass orifice to melt (R. 766-769 b), he subsequently directly contradicted himself. (R. 832 b) — 607 — In arguing this appeal the parties have all but ignored the aforementioned expert testimony relating to the condition of the Coleman Heater after the fire. This is particularly surprising in that if the jurors chose to give credence to Professor Young’s expert testimony in this regard, they could well have found that at some point an excessive amount of propane gas reached the heating unit. If this could not have occurred as a result of a fire of outside origin, it would necessarily follow that it occurred as a result of a failure of either the RX System or the Coleman Heater. The evidence indicates that the RX System at Marineland contained a number of built-in safety features. Among these features were fusible plugs placed at the top of each cylinder. According to the testimony of the defendants’ experts, the plugs were designed to melt when exposed to a temperature of approximately 165 degrees fahrenheit. (TR 786-788 b) From the testimony of these experts, the jury could well conclude that if the cylinders had been exposed to a fire of outside origin the plugs would have melted at 165 degrees fahrenheit and the excessive pressure caused by the heat would have been dissipated through the tops of the cylinders rather than through the regulator which led to the Coleman Heater.3 If this had occurred, there would obviously have been no indication after the fire that the bottom of the Coleman Heater had been subjected to the high temperatures which would have been generated if an excessive amount of fuel had reached the heater. Viewed in the abstract, the evidence discussed thus far would clearly support a jury finding that the fire at Marine-land resulted from a failure of the RX System. It was plaintiffs’ position in the District Court that such failure of the RX System was caused by the effects, over a period of time, of either one or a combination of the following items of negligence or defects in the design and' installation of the system: (1) The failure of the defendants to cover the manifold caused the manifold to become colder than the propane gas that goes into it, and liquid propane was thereby allowed to sit in the inlet valve of the regulator and then enter the regulator itself causing: (a) an increase in pressure of gas at the appliance, (b) freezing of the regulator because on entering the regulator the liquid propane expands and its temperature drops to minus 45 degrees fahrenheit, and (c) deterioration of the regulator valve seats and diaphram because of the minus 45 degree temperature. (2) In installing the regulator on the RX System a stripped thread fell into the regulator and over a period of time caused the same to malfunction. (3) The valve seat in the regulator was dented at the time of the fire, and this fact could have caused the fire by allowing a pressure increase at the appliance. (4) The RX System violated the Liquid Petroleum Gases Regulations of the state of Michigan in that: (a) the manifold was not pitched back in such a manner that condensation would drain back into the container, and no suitable means was provided for the vaporization of the condensation, (c) the regulator was not approved by Underwriters Laboratories for use in a system which allowed liquid propane to enter the regulator, and (d) the liquid level of propane in the tanks exceeded the allowable limit. — 608 — (5) An inappropriate regulator was placed on the RX System at Marine-land. (6) The installation of the cylinders of the RX System next to a wooden wall opposite a furnace caused a difference in temperature between the cylinders nearest the wall and those farthest from the wall, increasing condensation and the amount of liquid in the manifold and on the inlet valve of the regulator. To support these claims, the plaintiffs relied for all practical purposes solely upon the testimony of Professor Young. As previously noted, it is the defendants’ position that Professor Young’s opinion testimony relative to the causes of the failure of the RX System was predicated solely upon facts or inferences which by uncontradicted reasonable evidence were shown not to exist. Defendants urge, therefore, that Professor Young’s testimony in that regard must be entirely disregarded. It is notable that no similar claim is made with regard to that portion of Professor Young’s opinion testimony relative to the physical condition of the Coleman Heater after the fire. To support their position that Professor Young’s opinion testimony relative to the causes of the failure of the RX System must be entirely disregarded, defendants rely heavily upon the results of certain tests conducted by one of their experts, Robert P. Berman. Recognizing that Professor Young’s conclusions were predicated upon a view that the manner in which the RX System was designed and installed allowed a significant amount of liquid propane to accumulate above the inlet valve pad of the regulator, defendants attempted to demonstrate by way of Mr. Berman’s tests that such a result would not have occurred under the conditions that existed at Marineland on the morning of the fire. If it were to be concluded that the Berman tests substantially duplicated the conditions that existed at Marineland on the morning in question, the results of the same would, for all practical purposes, prove Professor Young’s theory to be invalid. This would follow for the reason that under Berman’s tests an insignificant amount of condensate accumulated at the inlet valve pad of the regulator and, further, the regulator did not permit an excessive amount of propane gas to pass through to the appliance. Plaintiffs, who conducted no tests, urge, however, that the evidence was such that the jury could properly conclude that the Berman tests were not performed under conditions which substantially duplicated those that existed at Marineland on the morning in question. It was Professor Young’s opinion that condensation would form in the manifold of the System primarily as a result of the transfer of heat from the wall of the building to the RX System. In testifying with regard to the effects of such heat transfer, Professor Young stated that with slight temperature differentials — such as a degree or so — the System would be thrown out of equilibrium and condensate would form in the manifold. Defendants vigorously urge that by this testimony Professor Young was indicating that the maximum temperature differential on the morning in question was approximately two degrees. Based upon such an interpretation of Professor Young’s testimony, defendants urge that the Berman tests, which were conducted with a fifteen degree temperature differential, conclusively established that there could not have been a substantial accumulation of condensate on the morning of the fire.4 — 609 — The interpretation of Professor Young’s testimony urged by the defendants is unduly restrictive. While Professor Young did speak in terms of a slight temperature differential destroying the equilibrium of the System, the jury could reasonably conclude from the whole of his testimony that he was speaking in terms of general principles rather than indicating the maximum temperature differential which existed on the morning of the fire. Thus, the jury could have concluded that Professor Young was merely expressing an opinion that under all of the circumstances, including such factors as heat transfer, atmospheric conditions, wind direction and velocity, and the manner in which the system was installed, there would have been a sufficient amount of condensation to cause a substantial accumulation of liquid propane in the manifold and at the inlet valve pad of the regulator. Therefore, we are unable to agree with defendants’ view that Berman’s tests conclusively rendered invalid Professor Young’s opinion that there would have been a substantial accumulation of condensate in the manifold and at the inlet valve seat of the regulator. Neither are we able to agree that the results of the Berman tests conclusively established that the factors taken into consideration by Professor Young could not have cooperated over a period of time to cause a failure of the RX System. As the defendants have observed, their experts disagreed not only with nearly every aspect of Professor Young’s opinion that defects in the design and installation caused an excessive amount of fuel to reach the Coleman Heater but, also, with nearly every underlying fact upon which Professor Young based his opinion. While some of these differences will be discussed in greater detail subsequently, it would be a work of supererogation to fully review in this opinion each and every point of difference. Manifestly, however, it was within the jury’s province to determine the credibility of the witnesses, whether they be experts or laymen. Defendants urge that the District Court erred in allowing the jury to consider the plaintiffs’ claim that a stripped thread could have caused or contributed to the failure of the regulator at Marineland. In support of this position, defendants claim that no evidence was offered from which the jury could reasonably conclude that a stripped thread at some point fell into the regulator. Orrin R. Hale, who installed the regulator at Marineland, testified, however, as follows: “Q. Well, is it a fact, sir, that you had difficulty with installing that regulator on the valve assembly in Exhibit 29 ? A. I could have. I don’t remember now. Q. Let’s see if I can refresh your recollection, sir. First, let me ask you this: Do you recall whether or not, in attempting to put that regulator on Exhibit 29, you stripped a thread on the fitting and you then put a different fitting on and that was the reason you had to run the pigtail because you did not have a right-hand thread fitting for the stub that copies down from Exhibit 29, do you recall that testimony? A. I could be. Q. Now, the stub that — or the portion of the regulator, sir, where the thread stripped would be this portion here, would it not? That would be the portion that would or would not have the left-hand thread or the POL fitting, is that correct? A. Yes.” (R. 311 b) With regard to Mr. Hale’s testimony, Professor Young stated as follows: “Q. Professor Young, there was some testimony in this matter that on Exhibit 28, which is the replacement regulator, the fitting immediately above setting C 256 on the regulator, that the threads had been stripped when this regulator was initially attempted to be installed on the bottom of the vertical pipe as shown in Exhibit 28. — 610 — Will you please tell us what happens when you strip a thread. A. May I have the other model ? Q. Yes. ****** A. This stripping a thread, a portion of the thread itself is sheared off the part. You literally tear off or break off part of the thread. Q. In this case, Dr. Young, would that portion of the thread that you break off be of what material? A. It would be of the softer material. Q. Would it be wood or metal ? A. It would be metal. Q. And from the installation from the arrangement of the RX System shown in Exhibit 20, and the photograph Exhibit 3, and the duplicate, I will call it by number, Exhibit 38, what would happen or what could happen to the portion of the metal from that stripped thread ? A. It would fall down into the regulator.” (R. 427 b) In view of Mr. Hale’s and Professor Young’s testimony, we find no error in the District Court’s allowing the jury to consider the plaintiffs’ “stripped thread” theory. The District Judge allowed the jury to consider whether the RX System installation at Marineland violated the Liquid Petroleum Gases Regulations of Michigan, and instructed the jury that violations of those regulations would be evidence of negligence. Defendants urge reversible error in this regard on the ground that the undisputed evidence showed the RX System installed in November of 1958 to have been approved by the Michigan fire marshal. The undisputed evidence relied on by the defendants was a “news flash” which the Pyro-fax Gas Corporation sent to all of its Pyrofax bulk distributors in the southeast Michigan area on July 18, 1960, nearly a year and a half after the installation of the RX System at Marineland. The “news flash” stated in pertinent part that: “The State Fire Marshal has finally approved the installation of the ‘Pyro-fax’ Gas RX Bulk Service.” Without passing upon the efficacy of defendants’ proof of “approval,” it must be noted that there was no evidence offered to indicate precisely what was “approved” by the State Fire Marshal and, further, there was no evidence offered to indicate that the installation at Marineland substantially conformed to whatever might have been approved by the fire marshal.5 The evidence offered did not require a finding that either the. specific installation at Marineland had been approved by the State Fire Marshal or that at the time of the fire the same conformed to whatever “system” might have been approved at some other time. Furthermore, evidence was admitted at the trial from which the jury could conclude that at the time of the fire certain aspects of the installation at Marineland did not conform to the Regulations. Under these circumstances, we conclude that it was proper for the District Judge to — 611 — allow the jury to consider the Regulations. See: Douglas v. Edgewater Park Company, 369 Mich. 320, 119 N.W.2d 567 (1963). As noted above, plaintiffs took the position in the District Court that the RX System at Marineland failed due to the effects over a period of time of one or a combination of many items of negligence or defects in the design and installation of the System. Faced with a general verdict, obviously we are unable to determine the specific basis upon which the jury found in the plaintiffs’ favor.6 However, from a thorough review of the entire record, and in accordance with our previous discussion, we conclude that no issues of fact were improperly submitted for the jury’s consideration. See: Roth v. Swanson, 145 F.2d 262 (C.A.8 1944). Moreover, we are of the view that the evidence, together with the legitimate inferences, was sufficient to support the general verdict in the plaintiffs’ favor. Sitta v. American Steel & Wire Div. of U. S. Steel Corp., 254 F.2d 12 (C.A.6 1958). Therefore, the District Judge properly denied defendants’ motions for a directed verdict and judgment notwithstanding the verdict. Defendants urge that the District Judge erred in failing to give certain of their requested instructions. Those instructions fall essentially within the following three categories: 1. Instructions pointing out the significance of certain items of evidence. 2. Instructions pointing out certain factual matters not in dispute. 3. A “burden of proof” charge in-v corporating and setting forth 17 of the plaintiffs’ factual claims. Further, defendants allude to 17 portions of the charge wherein they claim the District Judge erred. Defendants urge, in substance, that by reason of the District Judge’s failure to give their requested instructions and by his giving of certain instructions over their objection, the jury was misled with regard to the significance of certain items of evidence and also with regard to what the evidence reflected. As might well be gleaned from our discussion thus far, a multitude of facts were developed during the lengthy trial of this case. In his charge, the District Judge clearly and fairly summarized the factual claims of the parties and certain aspects of the evidence. Viewing the charge as a whole, we find no basis for the defendants’ claim that the District Judge misled the jury either with regard to the significance of certain items of evidence or with regard to what the evidence reflected. We conclude, therefore, that there was no prejudicial error in the charge as given or in the District Judge’s failure to give certain of the defendants’ requested instructions. Brown v. Addressograph-Multigraph Corp., 300 F.2d 280 (C.A.6 1962). In the middle of the trial, and after the plaintiffs had offered some evidence on the issue of damages, the parties met outside the presence of the court and reached an agreement as to the amount the plaintiffs would be entitled to recover in the event of a verdict in their favor. (R. 632 b-633 b) In colloquy between the court and counsel after this agreement was reached, no mention was made of the matter of interest from the date of the loss until the date of any potential judgment. Under Michigan law, the matter of interest from the date of a loss to the date of judgment is one which is properly submitted to the jury. Currie v. Fiting, 375 Mich. 440, 134 N.W.2d 611 (1965); Grand Trunk Western R. Co., v. H. W. Nelson Co., Inc., 116 F.2d 823 (C.A.6 1941). Notwithstanding this fact, in making their requests for instructions the plaintiffs included no request that the jury be instructed on the matter of interest from the date of the loss to the date of judgment. Rather, the plaintiffs’ requested instructions repeatedly stated that in the event of a verdict in — 612 — the plaintiffs’ favor, the jury was “to return a verdict for the plaintiffs in the amount stipulated against Jerry Ritter Appliances,, Inc., and Pyrofax Gas Corporation.” In his charge, the District Judge made no mention of the matter of interest from the date of the loss until the date of judgment. The plaintiffs took no exception to the charge in this regard. Approximately six months after judgment was entered against the defendants, the plaintiffs filed a “Motion for Interest” wherein they requested interest from the date of the loss until the date of the judgment. The District Judge denied this motion “for the reason the parties stipulated damages for which the judgment should be entered in the event the jury found liability.” Taking into consideration all of the circumstances mentioned above, we conclude that the plaintiffs’ Motion for Interest from the date of loss until the date of judgment was properly denied. The judgment of the District Court is affirmed. Footnotes [1] The plaintiffs also made a claim against the Coleman Company, manufacturers of the Coleman Wall Heater. On the basis of a jury finding in favor of Coleman, the District Court entered a judgment against the plaintiffs. No appeal has been taken from this portion of the District Court’s judgment. [2] Plaintiffs have raised certain questions by way of cross-appeal. In light of the conclusion which we ultimately reach, it will be unnecessary to consider or pass upon these questions. [3] It should be noted that the regulator on the RX System contained a rubber dia-phram. While this crucial part of the regulator could obviously be damaged by a certain degree of heat, there is evidence to indicate that the rubber used in this diaphram was compounded in such a manner that it would not be significantly damaged by that degree of heat needed to melt the fusible plug (R. 994, 995 b, 1034-1035 b), and that, therefore, if the fire had an outside origin the excessive gas pressure would have been released through the tops of the cylinders prior to any failure of the rubber diaphram. [4] It is clear that the Berman tests were not performed under conditions identical to those that existed at Marineland at the time of the fire. For example, while under the Berman tests the temperature of all of the cylinders was the same, there was evidence upon which the jury could very reasonably conclude that this would not have been the situation that existed at the time of the fire at Marineland. Moreover, there was some evidence tending to indicate that this difference might materially affect the validity of the Ber-man tests. [5] Defendants refer to a portion of the testimony of Mr. Ritter, President of defendant Jerry Ritter Appliances, Inc., in urging that the System as installed at Marineland was approved by the State Fire Marshal. With regard to this matter, Mr. Ritter stated as follows: Q. This RX System that you testified you installed in December of 1958 at Marineland, was that system approved by the State Fire Marshal, do you know? A. I don’t know the details. I understand specific instructions, specific authorization was given for that installation or a few installations. Q. You don’t know whether or not this system had been approved for installation by the State Fire Marshal? A. It is my understanding that the State Fire Marshal had given approval for a few installations. (R. 257 b) This testimony by Mr. Ritter is at best quite equivocable. Certainly it did not require the District Judge to withdraw the matter of the Liquid Petroleum Gases Regulations from the jury. [6] It should be noted that in this appeal there is no claim made that a distinction should be drawn between the liability of Pyrofax and that of Ritter.