TRANS AMERICA HOLDING, INC., Highland Village Holding, Inc., Westheimer National Bank, Josephine Dawson d/b/a Antiques of Europe a/k/a Glayva Antiques, and Smith & Armbuster d/b/a Gayles, Appellants, v. The MARKET-ANTIQUES AND HOME FURNISHINGS, INC., Appellee, 39 S.W.3d 640

OPINION ON MOTION FOR REHEARING LEE DUGGAN, Jr., J. (Assigned). We issued an opinion affirming the trial court’s judgment on August 17, 2000. Appellants filed a motion for rehearing. The motion for rehearing is denied. However, we withdraw our opinion of August 17, 2000, and issue the following opinion in its place. This is a negligence suit for property damage resulting from a shopping center fire. The trial court entered a take-nothing judgment following a jury verdict finding no negligence. Appellants assert two points of error on appeal. We affirm. I.

BACKGROUND Appellee, The Market-Antiques and Home Furnishings, Inc., opened a new store in Houston in the Highland Village shopping center in 1992. During the 1992 Christmas season, the store was decorated with seven artificial Christmas trees with tree lights, distributed in three rooms: the potpourri room, the narrow room, and the Christmas room. The trees remained lighted 24 hours a day during this season. On October 20, 1992, after appellee’s store was closed for the evening, a fire started in the store, damaging appellee’s facilities and the shopping center.

Appel-lee’s neighbors, two stores and a bank, were damaged by smoke, soot, and water from the fire department’s attempt to extinguish the fire. Appellants, the injured parties, sued (1) appellee, (2) the manufacturer of the Christmas light strings, and (3) the manufacturer of the tree.

Appellants alleged in their original petition that the Christmas tree lights “short circuited due to a design defect in the lights,” and that “due to a design defect in the Christmas tree, the tree caught on fire when the lights short circuited.” Appellants alleged that appellee was negligent for “creating, or otherwise allowing, a condition that would likely result in a fire”; that the tree manufacturer was negligent and strictly liable for designing and selling a Christmas tree that was not fire-proof or fire retardant; and that the Christmas light manufacturer was negligent and strictly liable for designing and selling unsafe lights.

Appellants dismissed the tree manufacturer in their first amended petition; after discovery, appellants settled with the manufacturer of the lights, Kurt S. Adler, for $337,500.00. In their third amended petition, appellants sought recovery from appellee alone, alleging it negligently caused the fire by placing too many lights on the trees, placing electric ornaments on the light strings, and leaving the lights on 24 hours a day. Appellants also alleged res ipsa loquitur.

At trial, appellants introduced expert testimony that the fire began in the Christmas tree room with the gold Christmas tree, caused by heat generated from the Christmas lights and/or a loose connection in the light sockets. For the court’s charge, appellants requested an instruction on res ipsa loquitur in connection with their negligence question. The trial court refused this request. The trial court entered judgment on the unanimous jury verdict finding that the fire was not caused by appellee’s negligence. Appellants assert two points of error on appeal. We affirm.

Appellants’ first point of error, a factual sufficiency challenge, urges that the jury’s finding was against the great weight and preponderance of the evidence. Appellants’ second point of error asserts the trial court committed reversible error when it refused to instruct the jury on res ipsa loquitur. II. DISCUSSION A. Point of error one — Whether the jury’s finding of no negligence was against the great weight and preponderance of the evidence 1. Whether appellants waived point of error one by failure to brief We first address a procedural argument raised by appellee.

The negligence question submitted to the jury encompassed both negligence and proximate cause — “Did the negligence, if any, of the parties listed below proximately cause the occurrence in question?” Appellants’ original brief generally challenges the jury’s finding of “no negligence” without expressly stating that the refusal to find proximate cause was against the great weight and preponderance of the evidence. Ap-pellee contends that appellants waived their factual sufficiency challenge by arguing the issue of negligence on appeal without addressing proximate causation.

Appellants respond that their challenge to the finding on negligence incorporated all of the elements of a negligence cause of action, not simply the breach of duty apart from causation. They cite Texas cases referring to negligence in this broad sense. See, e.g., Firestone Steel Prods. Co. v. Barajas, 927 S.W.2d 608, 613 (Tex.1996) (observing that “a negligence cause of action consists of: (1) a legal duty owed by one person to another; (2) a breach of that duty; and (3) damages proximately resulting from the breach”); Southerland v.

Kroger Co., 961 S.W.2d 471, 472 (Tex.App.—Houston [1st Dist.] 1997, no pet.) (“To establish negligence, evidence must be produced to establish a duty, a breach of that duty, and damages proximately caused by the breach.”) (emphasis added). Further, appellants expressly assert in their original brief that the evidence at trial established the fire was caused by appellee’s negligence. We find that appellants’ challenge addresses both the issues of negligence and proximate cause.

See Tex.R.App.P. 38.1(e) (“The statement of an issue or point will be treated as covering every subsidiary question that is fairly included.”). We hold that appellants did not waive their factual sufficiency argument for failure to address proximate causation. 2. Factual sufficiency analysis In reviewing a challenge to the factual sufficiency of a jury finding, Texas courts are “diligent in exercising restraint in the use of their power of conclusiveness over facts” out of long-established deference to jury verdicts. Herbert v. Herbert, 754 S.W.2d 141, 144 (Tex.1988).

Thus, we weigh all of the evidence and set aside the verdict “only if the evidence is so weak or the finding so against the great weight and preponderance of the evidence that it is clearly wrong and unjust.” Sisters of Charity of Incarnate Word v. Gobert, 992 S.W.2d 25, 28 (Tex.App.—Houston [1st Dist.] 1997, no pet.); see also Sims v. Washex Mach. Corp., 932 S.W.2d 559, 565 (Tex.App.—Houston [1st Dist.] 1995, no writ) (determining that a jury finding of no negligence “was not against the great weight and preponderance of the evidence”).

In arguing that the jury’s finding of no negligence hy appellee was against the great weight and preponderance of the evidence, appellants cite the testimony of two expert witnesses, a former City of Houston Fire Chief and an electrical engineer, who stated that the fire started with the gold Christmas tree in the Christmas room. Appellants introduced evidence that eight Sylvestri electric ornaments (sold by a different company than Adler) were put into light sockets on the Adler light strings on the gold Christmas tree.

Appellants’ witnesses opined that the cause of the fire was excessive heat in the gold Christmas tree. Appellants also introduced the testimony of appellee’s employees to show there were problems with excessively hot lights and an improper fit between the Sylvestri ornaments and the Adler light strand sockets. Appellee produced no experts of its own to rebut this evidence. However, appellee elicited testimony from appellant’s witnesses that tended to undercut appellants’ theory of proximate causation.

The jury heard evidence of four separate possible causes of the fire: (1) fire in the gold Christmas tree caused by cumulative heat, possibly (but not necessarily) including heat generated by the Sylvestri ornaments; (2) fire in the gold Christmas tree caused solely by the Sylvestri ornaments; (3) fire in the gold Christmas tree caused by the Adler lights; and (4) fire in the narrow room caused by a candle. 1. Fire in the gold Christmas tree caused by cumulative heat V.E. Rogers, former Fire Chief for the City of Houston, testified that the fire started with the gold Christmas tree.

He based his opinion on his interviews with firefighters who extinguished the fire, arson investigators who investigated the fire, photographs of the fire, and depositions and reports prepared by others. Rogers testified that the cumulative effect of heat from the number of lights on the tree, loose connections between the lights and sockets, and the higher voltage increase for the Sylvestri lights lowered the ignition temperature of the tree so that it caught fire.

Rogers also testified that, in his experience, there is an average of 80-100 Christmas tree fires per year in Houston, some of which involve fire retardant trees. He testified that it is generally negligent to leave Christmas tree lights on 24 hours per day, and that appellee was negligent in (1) putting 2,000 lights on a ten foot Christmas tree when the manufacturer recommended no more than 700 lights on a tree of that height; and (2) using electric light ornaments by a different manufacturer, Sylvestri, when the light string manufacturer, Adler, recommended using only Adler bulbs.

Appellants introduced the testimony of one of appellees’ Dallas employees, Rusty Turnipseed, who oversaw the decoration of the gold Christmas tree in the Houston store. Turnipseed stated that one of the Adler light strands became so hot that he burned himself when he touched it, but he did not report the incident because this was a common occurrence. Turnipseed’s testimony lends support to Rogers’ testimony that cumulative heat (rather than a particular defect in the lights or problem with light sockets) could have caused the fire. 2.

Fire in the gold Christmas tree caused by the Sylvestri ornaments Appellants also introduced the testimony of an electrical engineer, Judd Clayton, who opined from post-fire photographs of the Christmas tree lights and electrical cords that the fire began in the upper extremity of a Christmas tree.

Clayton testified that he did not “believe that the number of sets [of Adler tree lights] on the tree was a factor in the ignition of this fire.” Instead, he believed the fire was caused by a bad connection between a Sylvestri electric ornament and an Adler light strand socket which generated excessive heat and set the tree on fire. Clayton based this opinion on the fact that, although the lights had been used continuously without a problem from October 5 until the fire on October 20, the condition of the lights changed on the Monday or Friday before the Monday fire because the Sylvestri ornaments were added.

Thus, engineer Clayton’s theory of proximate causation, that a loose connection between the Adler light strand socket and a Sylves-tri ornament was solely responsible for the fire, undercut Chief Rogers’ theory that the number of lights and their continuous use contributed to the cause of the fire. However, both Rogers and Clayton identified a loose connection or connections in the lights as a source of heat for the fire. Clayton also noted that the Adler lights contained a manufacturer’s warning on the box not to use other manufacturer’s lights with the light strand.

Clayton testified that (1) fire damage could have been prevented if it had occurred while appellee’s employees were present, and (2) the fire would not have started if appellee had followed Adler’s manufacturer’s warning. In questioning Clayton, appellants referred to the testimony of two of appellee’s employees that they encountered problems in attaching the Sylvestri ornaments because the fit was too loose for the sockets in the Adler light strings. 3.

Fire in the gold Christmas tree caused by the Adler lights Under cross-examination, Clayton admitted that a number of other problems with the Adler lights could have caused the fire. He testified that when he examined other Adler lights similar to those used on the gold Christmas tree, he noted manufacturing defects — “abnormalities in the wiring, in the sockets, in the contacts, [and] some frayed wiring.” However, he concluded that manufacturing defects such as in these other Adler lights did not cause the fire because the tree had been lighted without a problem from October 5-20, 1992.

Then, Sylvestri ornaments were added and the fire occurred within days. Based on this change of circumstance and testimony from appellee’s employee that some of the ornaments fit loosely in the Adler light sockets, Clayton concluded that the fire that occurred shortly thereafter was caused by the high-resistance connection when the Sylvestri ornaments were added. However, Clayton produced no studies indicating that loose connections between sockets and ornaments cause fees.

He admitted that he had never physically examined the model of Sylvestri ornament that had been added to the Adler lights, nor had he tested samples of the Adler lights or Sylvestri ornaments.

Clayton also admitted that a regular tree light bulb with a loose connection could cause a high resistance problem such as he attributed to the Sylvestri ornaments, and that “from the standpoint of a loose connection, there is no difference between that Sylvestri ornament and a bulb.” Under cross-examination, Clayton may have been impeached in the jury’s view when he read a statement from his deposition testimony that he had inspected Adler Christmas lights (such as those used on the gold Christmas tree), that the Adler lights contained manufacturing defects, and that “[a]ny of those [manufacturing defects] that would have resulted in a loose connection could possibly have given rise to a fire.” Although Clayton opined that a loose Sylvestri ornament connection was not only a possible but the probable cause of the fire, he conceded the following problems also were possible causes of the fire: 1. the lights being wrapped too tightly around the tree, such that the wire was damaged or suffered abrasion; 2. the Adler miniature bulbs being inserted into the sockets incorrectly; 8. the Adler miniature bulbs being loosened as a light string was moved around the tree; 4. a defect in the string wire; 5. a defect in the string socket; 6. a bad connection where the light strings plugged together (power connections); and 7.a bad connection where the light strings plugged into the extension cord.

Appellee bolstered Clayton’s concession, ie., that a problem with the Adler lights could have caused the fire, by reading to the jury appellants’ earlier allegations of design defects. Appellants’ first original petition had claimed, regarding the origin of the fire: [O]n or about October 20, 1992, after the employees of The Market closed the store and left for the day, a fire broke out, causing severe smoke and water damage to Plaintiffs property. Upon information and belief, the employees of The Market left certain Christmas tree lights on in the store.

Because of a design defect in the Christmas tree lights and/or the negligence of The Market in leaving the lights on, the lights short-circuited and overheated. This short-circuit and/or overheating of the lights caused a fire to break out. The Market has conducted its own investigation and concluded that the employees of The Market left the Christmas lights turned on. These lights short-circuited due to a design defect in the lights. The fire then spread from the Christmas tree to other parts of the store. (Emphasis added). 4.

Fire in the narrow room caused by a candle falling on furniture Appellants themselves elicited from Chief Rogers that a Houston Fire Department Report written by Investigator Duane Hurley concluded that a candle fell onto a sofa and set it on fire in the narrow room. Further, appellants’ (as a possible preemptive move) introduced before the jury evidence of Investigator Hurley’s original and supplemental investigative reports, Plaintiffs Exhibit No. 14.

In his original investigation report, Investigator Hurley concluded that the fire was “an accidental fire caused by a candle left burning which ignited a stuffed chair [described elsewhere in the report as a ' “couch”'] and spread throughout the store.” Hurley related the accounts of several witnesses who told him that candles had been distributed throughout the store, and noted one witnesses’s statement that appellee “had candles burning all over the store in small glass jars.” Another witness asked Hurley “whether it could have possibly been a candle that started the fire.” In his supplemental report, Hurley concluded that [Tjhere was no normal cause for the fire.

There was [sic] no electrical problems and there was [sic] no other problems because this investigator had checked out the electrical problems. Hurley also related in his report the conclusions of two other investigators who affirmed his analysis that the fire had started with a piece of furniture, rather than from an electrical problem. Appellee attempted to impeach the investigative report’s analysis with testimony from its own employees that candles were not kept in the narrow room and were never left lighted after the store closed for the day.

Evidence tending to show appellee was not negligent Some testimony tended to undercut the allegation that appellee acted negligently. As discussed above, Clayton conceded on cross-examination that the number of lights, and the amount of time they were left on, did not cause or contribute to the Christmas tree fire.

He further testified that, without either a manufacturing defect in the Adler lights, a loose connection between a light strand and a Sylvestri ornament, or something combustible near the lights, he did not “have a problem with leaving those lights on.” This directly undercut Chief Rogers’ theory that the fire was caused by too many lights left on for too long.

Clayton conceded that, although the Adler tree lights contained a warning not to use other lights not included with the strand, the Sylvestri ornament package contained the statement, “Easily plugs into light bulb socket on any standard miniature light string.” Clayton also conceded that he had not tested whether the Sylvestri ornaments fit Adler Christmas light strands, or whether the match of the ornament to the socket made a loose connection. He also acknowledged that photographs of other Adler Christmas lights in Dallas showed manufacturing defects in the light sockets.

Appellee elicited testimony from an Adler representative that the instructions on Adler Christmas light packages did not recommend three fight strands per tree, as appellants had suggested in questioning their witnesses, but rather stated that only three fight strands should be connected directly to each other at a time. Further, he testified that the manufacturer’s recommendation of no more than 700 fights for a tree the size of appellee’s trees was a guideline, not a safety recommendation.

Appellee elicited testimony from one of its employees describing the placement of the Sylvestri ornaments on the gold Christmas tree. She testified that a total of eight ornaments were placed on the tree, that she personally placed five of the eight on the top half of the tree (where Clayton testified the fire started), and that all five of the Sylvestri ornaments in the top half of the tree securely fit after she replaced three loose ones.

On cross-examination, however, she conceded that she was not watching whether another employee, who placed the three ornaments on the bottom half of the tree, had difficulty securing a tight fit. She also conceded that she had on occasion discovered a candle was still lighted as she was leaving the store. Conflicting theories of causation This evidence presented the jury with conflicting theories of causation. If the fire was caused by a candle in the narrow room, then none of the electric Christmas light causation theories accounted for the fire.

Alternatively, if the jury believed the fire was caused by heat from the Sylvestri ornaments in the top half of the gold Christmas tree, then the jury could have found, as Clayton concluded, that the cumulative effect of the number of lights and the length of time they were left on did not cause the fire. Further, the jury could have accepted engineer Clayton’s testimony that the fire started in the upper extremity of the gold Christmas tree, while rejecting his assertion that the cause was a loose connection with the Sylvestri ornaments.

This was plausible because of testimony by appel-lee’s employee that she personally made certain that all the lights in the top half of the tree had a tight connection. In other words, the jury could have concluded that, as Clayton conceded was possible, the fire was caused by a defect in the Adler tree lights — whether an abrasion in the wiring, a loose connection with the Adler miniature bulbs, a problem with the connection where the lights plugged together, a manufacturing defect, or a design defect as appellants originally alleged.

Moreover, if the jury concluded that the fire was caused by a defect in the lights or wiring, about which appellee’s employees were unaware, the jury could also have agreed with Clayton’s testimony that there is generally “no problem” with leaving on non-defective lights. That is, the jury could have concluded that the fire was caused by an unforeseeable defect in the lights unrelated to the manner in which appellee decorated the tree. Certainly, the state of the record was such that the jury could well have concluded appellee’s negligence caused the fire.

But it was the jury’s province, not ours, to judge the credibility of the evidence and resolve conflicts or inconsistencies. See Perseus, Inc. v. Canody, 995 S.W.2d 202, 205 (Tex.App.—San Antonio 1999, no pet.); GXG, Inc. v. Texacal Oil & Gas, 977 S.W.2d 403, 419 (Tex.App.—Corpus Christi 1998, pet. denied). To decide the jury’s verdict was against the overwhelming weight of the evidence would virtually require us to conclude as a matter of law that it is always negligent for a store owner to leave tree lights on after hours.

Under a factual sufficiency review, we may not “substitute [our] judgment for that of the jury, even if the evidence would clearly support a different result.” Maritime Overseas Corp. v. Ellis, 971 S.W.2d 402, 407 (Tex.1998). Here, 12 jurors unanimously found that appellee did not proximately cause the fire through negligence.

In light of Clayton’s concession (1) that he had observed in other Adler lights similar to those used on the Gold Christmas tree the presence of manufacturing defects — “abnormalities in the wiring, in the sockets, in the contacts, [and] some frayed wiring”; and (2) that a defect in the wiring, unrelated to the size or number of lights, could have caused the fire, we conclude the jury’s verdict was not so contrary to the overwhelming weight of the evidence that it was clearly wrong and unjust. See id. We overrule point of error one. B.

Point of error two — Whether the trial court committed reversible error in refusing to instruct the jury on the doctrine of res ipsa loquitur. Appellants assert the trial court abused its discretion in refusing their requested jury instruction on res ipsa loquitur to accompany the negligence question. Ap-pellee responds that a res ipsa loquitur instruction was not required for two reasons. First, appellants’ specific allegations of negligence and proximate cause precluded their use of the res ipsa loquitur.

Second, appellants did obtain a circumstantial evidence instruction, which precluded an additional instruction on res ipsa loquitur or, at the very least, prevented them from showing harm in its denial. We review the trial court’s decision on the submission of a jury instruction under an abuse of discretion standard. See Plainsman Trading Co. v. Crews, 898 S.W.2d 786, 791 (Tex.1995); Schorlemer v. Reyes, 974 S.W.2d 141, 144 (Tex.App.— San Antonio 1998, pet. denied).

In res ipsa loquitur cases, “[s]ound exercise of this discretion is particularly important ... to accommodate the wide range of situations which can arise in these cases.” Mobil Chem. Co. v. Bell, 517 S.W.2d 245, 256 (Tex.1974); Schorlemer, 974 S.W.2d at 145. A res ipsa loquitur instruction relieves the plaintiff of the burden of proving the defendant committed a specific act of negligence when it is impossible for the plaintiff to determine the sequence of events, or when the defendant has superior knowledge or means of information to determine the accident’s cause. Jones v. Tarrant Util.

Co., 638 S.W.2d 862, 865 (Tex.1982); Turbines, Inc. v. Dardis, 1 S.W.3d 726, 740 (Tex.App.—Amarillo 1999, no pet.). To be entitled to a res ipsa instruction, the plaintiff must show both (1) that the character of the injury is such that it could not have occurred without negligence; and (2) that the instrumentality causing the injury was under the sole management and control of the defendant. Haddock v.

Arnspiger, 793 S.W.2d 948, 950 (Tex.1990). “The first factor is necessary to support the inference of negligence and the second to support the inference that the defendant was the negligent party.” Turbines, Inc., 1 S.W.3d at 741; see also Mobil, 517 S.W.2d at 251. The plaintiff must also show that the likelihood of other causes is “so reduced that the jury can reasonably find by a preponderance of the evidence that the negligence, if any, lies at the defendant’s door.” Porterfield v. Brinegar, 719 S.W.2d 558, 559 (Tex.1986); Schorlemer, 974 S.W.2d at 145.

Here, because the fire arose in appellee’s facilities, the parties do not dispute the second factor, ie., that the instrumentality causing the fire was under the appellee’s sole management and control. But appellees argue that appellants failed to establish the first factor — that the fire was the sort of accident that occurs only with negligence. In analyzing this first factor, our focus is on the nature of the injury rather than the conduct of appellee.

See, e.g., Turbines, 1 S.W.3d at 741 (noting that “because application of the [res ipsa loquitur ] doctrine does not require evidence of a specific act of negligence, the focus is not on the ‘alleged negligent act’ but on the nature of the accident or injury”) (citing Prosser and Keeton on Torts, 5th ed. § 39 (West 1984)). Courts have generally found res ipsa inapplicable where (1) the plaintiff cannot rely on general knowledge to show that the accident would not occur without negligence, and (2) the plaintiff has not produced expert testimony that the injury ordinarily does not occur without negligence.

See, e.g., Soto v. Texas Indus. Inc., 820 S.W.2d 217, 220 (Tex.App.—Fort Worth 1991, no writ) (holding the trial court did not abuse its discretion in denying a res ipsa loquitur instruction when the plaintiffs “presented no expert knowledge that concrete walls do not ordinarily fall in the absence of negligence” and general knowledge could not support the contention). None of appellants’ experts testified that a fire such as occurred here cannot begin without negligence. Neither have appellants shown that there is general knowledge that such a fire cannot occur without negligence.

Appellants’ experts testified that specific conduct by appellee’s employees — acts about which there was direct evidence — caused the fire. This evidence, if true, tends to establish negligence directly. Texas courts have found that res ipsa loquitur is not appropriate when direct evidence points to possible negligence. See Fair v. Wright, 833 S.W.2d 597, 600-01 (Tex.App.—Corpus Christi 1992, writ denied) (determining that res ipsa loqui-tur, “which concerns cases involving only weak circumstantial evidence,” was inapplicable when both circumstantial and direct evidence pointed to a specific act of negligence).

In short, though appellants produced direct evidence possibly indicative of negligence, they did not establish that the character of the fire was such that, taken apart from appellee’s alleged acts, negligence may be inferred. In any event, the Texas Supreme Court has indicated that a circumstantial evidence instruction may suffice in place of a res ipsa instruction: [i]n some cases the inferences arising from the circumstances of the accident are so apparent that no special instruction is necessary.

In other cases it is sufficient to give the jury a circumstantial evidence instruction to instruct them that acts of negligence can be proved both by direct evidence and by inferences from other facts proved. Finally, there are some cases in which it is helpful to give the jury a more specialized Res ipsa instruction to the effect that if they find the two required factors, they are entitled, but not compelled, to infer that the defendant was negligent. Mobil, 517 S.W.2d at 256 (emphasis added). Here, the jury was instructed that it could infer negligence from circumstantial as well as direct evidence.

Appellee argues (1) that Mobil prohibits the inclusion of both a circumstantial evidence and a res ipsa instruction in the same negligence question, and (2) that appellants cannot show harm because the final jury instruction instructed the jury on circumstantial evidence. We agree with appel-lee’s second argument; under these facts, appellants cannot show they were harmed by the inclusion of a circumstantial evidence inference instruction rather than an instruction on res ipsa loquitur.

Appellants were not deprived of a theory of recovery but merely limited to one means of instructing the jury on the inference of negligence. See, e.g., Martz v. Weyerhaeuser Co., 965 S.W.2d 584, 589 n. 3 (Tex.App.—Eastland 1998, no pet.) (observing that the doctrine of res ipsa loqui-tur does not provide an independent ground of recovery but, rather, is a rule of evidence which allows the inference of negligence in certain cases); Lucas v. Titus County Hosp.

Dist./Titus County Memorial Hosp., 964 S.W.2d 144, 155 (Tex.App.— Texarkana 1998, pet. denied) (Res ipsa loquitur “is not a separate cause of action from negligence; rather, it is simply a rule of evidence by which negligence may be inferred by the jury.”), pet. denied per curiam, 988 S.W.2d 740 (Tex.1998). Even if appellants had shown that the character of the fire was such that it could not have occurred without negligence, they have not shown that the error resulted in an improper judgment. See Tex.R.App.P. 44.1(a)(1); Island Recreational Dev. Corp. v. Republic of Tex. Sav.

Ass’n, 710 S.W.2d 551, 555 (Tex.1986) (holding that failure to issue appropriate instructions for the jury is not reversible error per se, but rather must amount “to such a denial of the rights of the complaining party as was reasonably calculated and probably did cause the rendition of an improper judgment”). Appellants have failed to show the trial court abused its discretion in refusing a res ipsa loquitur instruction to the jury. We overrule point of error two. We affirm the judgment. Justice O’CONNOR, dissenting. .

Appellants also introduced the testimony of expert witness James Stallcupp, the chief plans examiner for Fort Worth, Texas. Stall-cupp's position as chief examiner included the duty of ascertaining whether businesses and retail establishments comply with applicable local and national electrical codes. Stallcupp opined that appellee’s failure to follow applicable electrical codes caused the fire. However, he was unable upon cross-examination to cite a single local or national code

section existing in 1992 that expressly required that Christmas lights be turned off when the store was closed or specified a maximum number of Christmas lights per tree, though he contended that appellee was required to follow manufacturer recommendations for the lights. He conceded that the recommendations for the Sylvestri ornament specified that it could be used with miniature light strands. However, he stated that appel-lee should not have used the Sylvestri ornaments on the Adler light strands because the Adler light manufacturer did not specify the use of Sylvestri ornaments.

Stallcupp’s testimony tended to show the breach of a duty rather than proximate causation. He did not offer his own conclusions on the actual source of the fire but relied instead on the other experts’ analyses (that defects in the lights or loose connections sparked the fire) to conclude that appellee’s following the applicable code restrictions would have prevented the fire. . Throughout this opinion, "lights” or "bulbs” shall refer to the Adler light string miniature bulbs, while "ornaments” shall refer to the Sylvestri electric ornaments. .

Appellants state in their brief that appellee “stipulated” that the fire began in the Christmas tree room and assert that Hurley’s reports were never entered into evidence. However, they concede that testimony regarding the content of the investigative report was introduced at trial.

But appellee points out that appellants (1) refused to accept appellee’s offer to stipulate, and (2) offered evidence of the Fire Department investigative report’s conclusion that the fire had been caused by a lighted candle, both by eliciting testimony from Chief Rogers and by offering Hurley’s supplemental and original reports as Plaintiff's Exhibit No. 14. . Appellants’ argument on this point could be summed up as follows:

(1) Res ipsa is appropriate when it is hard to prove the cause of an accident; (2) here, appellee was clearly negligent in dealing with the Christmas lights, and but for this negligence the fire would not have occurred; (3) therefore, appellant should not have to prove the cause of the fire. But arguing that appellee negligently caused the fire does not establish that the fire could not occur in the absence of negligence.