OPINION ON REHEARING
LEE DUGGAN, Jr., Justice (Retired).
We deny the Motion for Rehearing filed by appellees Phoenix Assurance PLC, The Ocean Marine Insurance Company, LTD., and Compagnie D’Assurances Maritime, Aerine Et Terrestres. However, we withdraw our opinion dated May 10, 2001, and we issue this opinion in its stead.
Appellant, Mescalero Energy, Inc. (Mes-calero), appeals a summary judgment in favor of appellees, its insurers and underwriters, following an underground oil well blowout during drilling operations. Appel-lees denied coverage, contending that the explosion occurred within a single “formation,” an excludable event under the policy’s conditions of coverage. Because of the conflicting evidence presented, we find that the trial court erred in determining that the term “formation” was unambiguous and in granting summary judgment on that basis. We reverse and remand to the trial court.
I.
Background
In 1993, Mescalero, an oil and gas drilling contractor, began a program to drill [316]*316several horizontal wells in Fayette County, Texas. One of the wells, the Stork # 1 Well, was to be drilled horizontally through a previously existing vertical well. The location of the drilling is commonly referred to as the Austin Chalk.
Mescalero sought coverage for its program through an insurance agency, Carter & Company, and obtained a Blowout Insurance Policy (the “policy”) from the two appellee insurers.1 -The four other appel-lees underwrote the policy.2
The policy provided coverage for described costs in regaining control of a blowout, redrilling expenses, and third-party equipment damaged by covered casualties in Mescalero’s control. A policy endorsement governing underground blowouts states that “[a]n insured well shall also be deemed to be a blowout when there is a sudden, accidental, uncontrollable and continuous flow of oil, gas or water simultaneously between two or more separate formations via by [sic ] the well bore under the surface of the earth or water bottom.”3 (Emphasis added.)
On March 23, 1994, an uncontrolled flow of oil, gas, or water occurred in the Stork # 1 Well. This caused a loss of normal fluid circulation in the well bore and trapped down-hole equipment Mescalero had leased from Horizon Directional Systems, Inc. After several unsuccessful retrieval attempts, Mescalero cut the drill pipe and abandoned the down-hole equipment. When Horizon sued Mescalero for the value of the abandoned equipment, Mescalero forwarded the petition to its insurers for defense under the policy.
The insurers retained Riseden Services, Inc., an oil and gas engineering consulting firm, to analyze the technical aspects of the March 23rd incident. J.E. Riseden, a petroleum engineer and president of Rise-den Services, prepared a report concluding that gas and fluid moving through a fractured, unstable portion of the single formation through which Mescalero had drilled, the Austin Chalk, caused the formation to cave in on the drill pipe. Riseden testified in his deposition and his summary judgment affidavit that he found no evidence that oil, gas, or water had flowed from one formation through the well bore to another separate formation, and concluded that there was no evidence of a “blowout” as defined in the policy. The insurers and underwriters denied the occurrence of an underground blowout, denied coverage for the March 23rd incident, and refused to defend Mescalero against Horizon’s lawsuit.
The insurers and one of the underwriters, Planet Indemnity Company, then filed a declaratory judgment action seeking a determination that Mescalero’s damages were not covered under the terms of the policy because, inter alia, the damages resulted from a “kick,”4 and not from a “blowout” as defined in the policy. They asserted that the accident occurred within a single formation, the Austin Chalk, and not between separate formations, as re[317]*317quired to constitute an underground blowout. Mescalero answered and brought counterclaims against the insurers and underwriter Planet Indemnity Company, asserting that the Austin Chalk was a series of formations, and that the explosion and damage occurred between two or more of them. In an amended counterclaim, Mes-calero joined the remaining underwriters and the insurance agency, Carter & Company, as defendants.5
The insurers and underwriters moved for summary judgment on the ground that, because the Austin Chalk is a single formation, the uncontrollable loss of fluids did not occur “between two or more separate formations” (as required by the policy) and the underground blowout endorsement did not provide coverage for the March 23rd accident.6 The trial court granted the insurers’ and underwriters’ motions for summary judgment, severed Mescalero’s remaining claim against Carter & Company, the insurance agency, and made the summary judgment final and appealable.7
II.
Standard of Review
The party moving for Rule 166a(c) summary judgment has the burden of showing that no genuine issue of material fact exists and that it is entitled to judgment as a matter of law. Nixon v. Mr. Property Mgmt. Co., 690 S.W.2d 546, 548 (Tex.1985). In deciding whether a disputed material fact issue exists, the reviewing court will take as true all evidence favoring the nonmovant. Id. at 548-49. Every reasonable inference must be indulged in favor of the nonmovant and any doubts resolved in the nonmovant’s favor. Id.
III.
Discussion
In a single point of error, Mescalero asserts the trial court erred in granting summary judgment in favor of the insurers and underwriters (collectively, “appel-lees”). Under Mescalero’s definition, the [318]*318Austin Chalk constitutes a series of formations.
Mescalero argues that the term “formation” is ambiguous because it is subject to two reasonable definitions, that any ambiguity should be resolved in favor of the insured, and that because of this ambiguity, the trial court erred in granting summary judgment for appellees on the issue of coverage.
Our analysis turns on the meaning of the term “formation” and the evidence to which we may resort to make this determination.
A. Appellees’ Summary Judgment Motions
In their motions for summary judgment and on appeal, appellees cite Howaed R. Williams & Chaeles J. MeyeRS, Manual of Oil and Gas Terms (9th ed.1994), a commonly cited oil and gas dictionary, which defines a “formation” as
A succession of sedimentary beds that were deposited continuously and under the same general conditions. It may consist of one type of rock or of alterations of types. An individual bed or group of beds distinct in character from the rest of the formation and persisting over a large area is called a “member” of the formation. Formations are usually named for the town or area in which they were first recognized and described, often at a place where the formation outcrops. For example, the Austin chalk formation outcrops at Austin, Texas.
Id. at 559 (emphasis added). Appellees note this definition of “formation” was quoted in its entirety (including the reference to the Austin Chalk) by the Texas Supreme Court in Amarillo Oil Co. v. Energy-Agri Products, Inc., 794 S.W.2d 20, 23 n. 3 (Tex.1990). They argue that the Court expressly adopted this definition, under which the Austin Chalk constitutes a single formation, as the sole and unambiguous meaning of the term within the oil and gas industry.
Appellees also cite the affidavit of J.E. Riseden (included with their motions for summary judgment). Riseden concluded that, based on his investigation, there had not been an underground blowout within the definition of the policy. In his affidavit, he also stated that the definition of “formation” in the Williams & Meyers dictionary is the one most generally accepted in the industry.
B. Mescalero’s Response
In its response, Mescalero argued that summary judgment was not proper because the term “formation” is subject to more than one reasonable interpretation. Mescalero attached the affidavit of Byron Davenport, a petroleum engineer whose resume and testimony show extensive experience with horizontal drilling techniques. Davenport concluded that the March 23rd accident was a blowout as defined in the policy and stated that, in his opinion, there are numerous formations within the Austin Chalk. According to Davenport, the ordinary, generally accepted meaning of the term “formation” for horizontal wells would be defined as “[a]ny mappable, separate, self contained pressure unit within a geologic interval.” However, Davenport did not cite any authority upon which he bases his definition.
Mescalero also contended that the deposition testimony of Riseden and of Jack E. Carter, a partner of the insurance agency that obtained the policy for Mescalero, establishes that the term “formation” is subject to more than one reasonable interpretation.
Based on Davenport’s affidavit, and the deposition testimony of Riseden and Carter, Mescalero argued that the trial court [319]*319was presented with two reasonable interpretations of the term “formation,” and was, thus, required to adopt the interpretation most favorable to the rights of the insured. At a minimum, Mescalero argued, its alternative interpretation of the term “formation” gave rise to a fact issue concerning whether an underground blowout had occurred under the policy.
C. Rules of Construction
Insurance policies are controlled by rules of interpretation and construction applicable to contracts generally. Forbau v. Aetna Life Ins., Co., 876 S.W.2d 132,133 (Tex.1994). For insurance policies in particular, however, we construe the policy against the insurer when ambiguous policy terms permit more than one interpretation, especially when the policy terms exclude or limit coverage. See State Farm Fire and Cas. Co. v. Vaughan, 968 S.W.2d 931, 933 (Tex.1998); National Union Fire Ins. Co. v. Hudson Energy Co., Inc., 811 S.W.2d 552, 555 (Tex.1991). Furthermore, the burden lies with the insurer to prove that an exclusion or limitation on coverage applies. Tex. Ins.Code Ann. art. 21.58(b) (Vernon Supp.2001).
The primary concern of a court in construing a written contract is to ascertain the true intent of the parties as expressed in the instrument. National Union Fire Ins. Co. v. CBI Indus., Inc., 907 S.W.2d 517, 520 (Tex.1995). If a written contract is so worded that it can be given a definite or certain legal meaning, then it is not ambiguous. Id.; Coker v. Coker, 650 S.W.2d 391, 393 (Tex.1983). Parol evidence is not admissible for the purpose of creating an ambiguity. See CBI Indus., Inc., 907 S.W.2d at 520; Universal C.I.T. Credit Corp. v. Daniel, 150 Tex. 513, 243 S.W.2d 154,157 (1951).
If, however, the language of a policy or contract is subject to two or more reasonable interpretations, it is said to be ambiguous. CBI Indus., Inc., 907 S.W.2d at 520; see Glover v. National Ins. Underwriters, 545 S.W.2d 755, 761 (Tex.1977). Whether a contract is ambiguous is a question of law for the court to decide by looking at the contract as a whole in light of the circumstances present at the time the contract was executed. See CBI Indus., Inc., 907 S.W.2d at 520; Coker, 650 S.W.2d at 394. Only where a contract is first determined to be ambiguous may the courts consider the parties’ interpretation and admit extraneous evidence to determine the true meaning of the instrument. CBI Indus., Inc., 907 S.W.2d at 520.
An ambiguity in a contract may be either “patent” or “latent.” Id. A patent ambiguity is evident on the face of the contract. Id. If a contract which is unambiguous on its face is applied to the underlying subject matter of the contract and an ambiguity appears by reason of some collateral matter, the ambiguity is latent. Sidelnik v. American States Ins. Co., 914 S.W.2d 689, 691 (Tex.App. — Austin 1996, writ denied).
D. Whether the term “formation” is ambiguous
Mescalero contends that a fact issue exists because the term “formation” has multiple, reasonable meanings when applied to the Austin Chalk and is therefore latently ambiguous. However, the claim that a contract term, on its face, is subject to two or more reasonable interpretations is an argument of patent ambiguity.8 See CBI Indus., Inc., 907 S.W.2d [320]*320at 520. Thus, we disagree with Meseale-ro’s characterization of its argument as one based on latent ambiguity, and consider this case based on the law applicable to patent ambiguity in the contract.
1. Use of extrinsic evidence to determine a term’s generally accepted meaning
A term not specifically defined by an insurance policy must be given its plain, ordinary and generally accepted meaning, unless consideration of the policy itself shows it to have been used in a different sense. See Ohio Cas. Group, of Ins. Companies v. Chavez, 942 S.W.2d 654, 658-59 (TexApp. — Houston [14 Dist.] 1997, writ denied). Courts may not look to extrinsic evidence to prove the existence of an ambiguity. See Friendswood Development Co. v. McDade + Co., 926 S.W.2d 280, 283 (Tex.1996) (“Only after a contract is found to be ambiguous may parol evidence be admitted for the purpose of ascertaining the true intentions of the parties expressed in the contract”); Sears, Roebuck and Co. v. Commercial Union Ins. Corp., 982 S.W.2d 151, 154 (Tex. App. — Houston [1st Dist .] 1998, no pet.).
Nevertheless, extrinsic evidence may “be admissible to give the words of a contract a meaning consistent with that to which they are reasonably susceptible, i.e., to ‘interpret’ contractual terms.” CBI Indus., Inc., 907 S.W.2d at 521. Texas courts often resort to the use of external references such as dictionaries to determine an insurance policy term’s plain, ordinary, and generally accepted meaning. See Heritage Resources, Inc. v. Nations-Bank, 939 S.W.2d 118, 121-22 (Tex.1996); Ramsay v. Maryland American General Ins. Co., 533 S.W.2d 344, 346 (Tex.1976).
In particular, a specialized industry or trade term may require extrinsic evidence of the commonly understood meaning of the term within a particular industry. See CBI Indus., Inc., 907 S.W.2d at 521 n. 6 (noting that extrinsic evidence may be consulted in determining the commonly understood meaning of terms within a particular “place, vocation, trade, or industry”).
2. Extrinsic evidence provided by the parties to show the generally accepted meaning of a “formation”
The Williams & Meyers definition
Appellees relied upon the Williams & Meyers definition of “formation” and the Supreme Court’s recognition of that definition in Amarillo Oil to show the trial court that the term “formation” is unambiguous and that the Austin Chalk is a single formation. See Amarillo, 794 S.W.2d at 23 n. 4 (defining the term “formation”).9 The Williams & Meyers definition of a “formation” is certainly one reasonable interpretation; however, Mescalero’s appeal turns on whether it is the only reasonable interpretation within the meaning of this blowout insurance policy.
Mescalero asserts that the term formation is subject to more than one reasonable interpretation, citing (1) the affidavit of its expert, Davenport, to establish a second reasonable interpretation, and (2) the testimony of appellees’ expert, Riseden, and of Jack E. Carter, a partner of the insurance [321]*321agency that obtained the policy for Mes-calero.
The Davenport affidavit
Mescalero attached the affidavit of Byron Davenport with its response to appel-lees’ motions for summary judgment. Davenport’s affidavit testified to his credentials, including a Ph.D. in petroleum engineering, his experience as a practicing petroleum engineer since 1984, and his curriculum vitae reflecting extensive experience in horizontal drilling. Davenport stated in his affidavit that he had reviewed well data for the Stork # 1 Well, the reports that Riseden Services, Inc. generated after the March 23rd accident, and the summary judgment evidence of the parties. Based on this information, Davenport opined that “there is conclusive evidence that an underground blowout did occur on or about March 23, 1994.” He added that “[t]he underground blowout which occurred was a sudden, accidental, uncontrollable and continuous flow of oil, gas or water simultaneously between two or more separate formations via the well bore under the surface of the earth or water bottom.” (Emphasis added.) He also stated that “there are numerous formations within the 'Austin Chalk” and that “[t]he ordinary, most easily understood and generally accepted meaning of the term ‘formation’ ” for horizontal wells would be defined as: “any mappable, separate, self contained pressure unit within a geologic interval.”
The record does not reflect that Davenport gave a deposition; therefore, the only evidence from him that we may consider is the affidavit.
Riseden’s testimony
Although Riseden testified that Davenport’s definition of a formation was not generally accepted and that Williams & Meyers’ definition was the term’s generally understood meaning, he conceded that he had never used the Williams & Meyers definition before his expert affidavit:
Counsel: You gave a definition of formation in your affidavit, correct?
Riseden: Yes.
Counsel: Have you ever used that definition in defining formation before?
Riseden: No.
Counsel: You would agree with me there’s more than one definition for formation, isn’t there?
Riseden: I’m sure there is. There could be.
Counsel: Would you agree with me that another acceptable definition of formation would be any mappable separate self-contained pressure unit within a geological interval?
Riseden: I would not support that definition .... I don’t know that it has industry-wide acceptance.... As far as I know, it does not. I don’t— without support, it is not accepted on the industry basis by some authority.
Counsel: Have you ever heard of anyone using your definition before of formation?
Riseden: Yes.
Counsel: Who?
Riseden: Similar parts, I mean, in context. I don’t know formally that, “Here is the definition of a formation.” I mean, in context that’s what’s generally understood to be a formation.
(Emphasis added.)
Riseden later reiterated that he was unfamiliar with Davenport’s definition as “a commonly accepted definition of formation,” and added that a commonly accepted definition is one that is “widely used” or “published by a reputable source.”
[322]*322Carter’s testimony
After Mescalero’s counsel questioned Carter regarding whether he personally believed an underground blowout had occurred within the meaning of the policy, Carter answered as follows:
Carter: If you would ask a layman if the blowout occurred entirely within the chalk, he would probably say, and this is speculation on my part, that it didn’t penetrate from one formation to another. But a street opinion isn’t worth a dime, you know, when it comes to the actual engineering facts. The expert may think of the — the chalk as being several formations. But there’s chalk and Eagleford and Edwards and so forth, and far be it from me to know if there are several different subdivisions within a formation. That would be different zones, and, you know, it could penetrate from one zone to another.
Counsel: And certainly since the policy doesn’t define formation, you can’t tell from the policy, can you?
Carter: I wouldn’t be able to.
Although the Davenport affidavit expressly states an alternate definition for a formation, we find that the testimony of Riseden and Carter does not establish the existence of an alternate definition substantively different from the Williams & Meyers definition. Riseden testified that Davenport’s definition is not accepted as an industry-wide commonly understood definition of formation, while the Williams & Meyers definition is.10 Carter’s testimony, on the other hand, professed agnosticism of the meaning of formation — merely establishing that he was not qualified to comment on an expert definition of formation.
Other than these witnesses, Mescalero has not cited to us any testimony (expert or otherwise) to support its proposed definition of a formation. Thus, having determined that the Williams & Meyers definition establishes one reasonable definition for a formation, our analysis turns on whether the Davenport affidavit alone is enough to raise a fact issue as to a second reasonable definition. This issue first requires us to determine whether Davenport’s testimony is admissible.11
3. Whether Davenport’s testimony is admissible as extrinsic evidence of the commonly understood meaning of a “formation”
Appellees contend their summary judgment motions were based solely on the Williams & Meyers definition and the express adoption of this definition in Amarillo Oil. They argue that Davenport’s affidavit should be excluded as impermissible parol evidence. We note, however, that appellees’ motions each cited Riseden’s testimony, an extrinsic source which used [323]*323the Williams & Meyers definition. Furthermore, we disagree with the appellees’ contention that Amarillo Oil expressly adopted the Williams & Meyers definition of a “formation” as the only and unambiguous meaning of the term. The definition of a “formation” was not central to Amarillo Oil ⅛ holding, and the Texas Supreme Court’s use of that definition did not establish that a formation is subject to only one commonly understood meaning within the oil and gas industry.
The principal issue in Amarillo Oil was the definition of “casinghead gas.” See 794 S.W.2d at 22. In answering that question, the Court referred to a variety of extrinsic sources, including Texas statutory code provisions and the Williams & Meyers dictionary. See id. at 22-25. In order to determine the meaning of casing-head gas, the Court referred in turn to the definition of a “stratum,” and noted that a stratum “may be the same as a ‘formation’ if there is only one type and layer of rock that was deposited continuously and under the same general conditions, or it may be a part (one layer) of a formation.” Id. at 23. The Court added the Williams & Meyers definition of a “formation” in a footnote, noting that “[a] ‘formation’ has been defined and explained in these terms.” Id. at 23 n. 3 (emphasis added). The Court did not hold that the Williams & Meyers definition is the only commonly understood meaning of a formation, and the opinion’s reference to this definition is mere dicta.
Amanllo Oil stands for one proposition, however, that is central to our case — courts may refer to extrinsic evidence such as industry dictionaries or statutory definitions to determine the commonly understood meaning of an industry term. Appellees concede (indeed, urge) that we refer to one form of extrinsic evidence — the Williams & Meyers dictionary. The question for us is: “Can a reasonable definition of an industry term be established through expert testimony?” We hold that it can.
We find no Texas cases holding that expert testimony is per se inadmissible to determine the common meaning of a specialized industry term within that industry. An industry resource such as the Williams & Meyers dictionary simply contains the definitions approved of, and asserted by, the individuals who compiled the dictionary. The principal differences between a dictionary definition and one offered by a current oil and gas expert in an affidavit may be no more than the age and the hard cover binding on the book.
Many courts have used expert definitions to determine the meaning of specialized terms before deciding whether an instrument is ambiguous. See, e.g., Construction Interior Sys., Inc. v. Mamott Family Restaurants, Inc., 984 F.2d 749, 755 (6th Cir.1993) (noting that under Ohio law, “contract terms are to be given their plain, ordinary meaning unless ... expert testimony provides a different meaning, particular to the industry involved ”) (emphasis added); Carondelet Sav. & Loan Ass’n v. Citizens Sav. & Loan Ass’n, 604 F.2d 464, 470 (7th Cir. 1979) (discussing the defendant’s use of expert testimony and learned treatises to prove the meaning of a term); see also Markman v. Westview Instruments, Inc., 52 F.3d 967, 986 (Fed.Cir.1995) (noting in a patent claim dispute that a court may consider extrinsic evidence to help it interpret the meaning of technical terms as they are understood by those familiar with the subject, and that it is the court’s unfamiliarity with the terminology of the technical field rather than ambiguity in the patent claim that permits the use of such extrinsic evidence), affd, 517 U.S. [324]*324370, 116 S.Ct. 1384, 134 L.Ed.2d 577 (1996).
We hold that Davenport’s expert affidavit, offered by Mescalero to show the commonly understood meaning of a formation, is not inadmissible for that purpose simply because it is extrinsic to the policy.
4. Whether the Davenport affidavit’s definition is conclusory
Appellees also argue that, assuming Davenport’s affidavit is not inadmissible as parol evidence, the affidavit is not competent summary judgment evidence because it merely states Davenport’s conclusions without supporting reasoning or other authorities on which he relied. Texas Rule of Civil Procedure 166a(f) requires that “supporting and opposing affidavits shall be made on personal knowledge, shall set forth such facts as would be admissible in evidence, and shall show affirmatively that the affiant is competent to testify to the matters stated therein.” Tex.R. Civ. P. 166a(f). Texas decisions have found expert witness testimony incompetent for summary judgment purposes when it does not adequately set out the basis for the opinion. See, e.g., Ryland Group, Inc. v. Hood, 924 S.W.2d 120,122 (Tex.1996) (“An interested witness’ affidavit which recites that the affiant ‘estimates,’ or ‘believes’ certain facts to be true will not support summary judgment.”); Anderson v. Snider, 808 S.W.2d 54, 55 (Tex.1991) (observing that “conclusory statements made by an expert witness are insufficient to supT port summary judgment”); Jensen Const. Co. v. Dallas Cty., 920 S.W.2d 761, 768 (Tex.App. — Dallas 1996, writ denied) (“An expert’s opinion is conclusory and will not support summary judgment if it does not contain the basis or reasoning for the opinion”).
In considering whether Davenport’s affidavit is merely conclusory, in violation of Rule 166a(f), we divide it into three major assertions: (1) that an underground blowout occurred between two or more separate formations on or about March 23, 1994; (2) that the Austin Chalk contains numerous formations; and (3) that “the ordinary, most easily understood and generally accepted meaning of the term ‘formation’ for horizontal wells” is “any mappable, separate, self contained pressure unit within a geologic interval.”
Cases holding an expert’s affidavit insufficient to support summary judgment have tended to turn on the expert’s omission of the basis or reasoning for his conclusion. See, e.g., Lara v. Tri-Coastal Contractors, Inc., 925 S.W.2d 277, 279 (Tex.App. — Corpus Christi 1996, no writ) (noting that “the experts failed to provide any legal basis or reasoning for deducing that [appellee] acted reasonably, took all necessary and reasonable precautions, and was not grossly negligent. Without this information, their testimony amounts to sworn denials of appellants’ claims.”); Hamlin v. Gutermuth, 909 S.W.2d 114, 117 (Tex.App. — Houston [14th Dist.] 1995, writ denied) (observing that, “[without sufficient reasoning or the basis for the affiants’ opinions that appel-lees caused the damages alleged by appellant, the affidavits are mere sworn denials, and, thus, incompetent to raise a fact issue and controvert a summary judgment.”). Here, in contrast, Davenport stated that his first two assertions — that (1) the blowout occurred between two or more separate formations and (2) the Austin Chalk contains many formations — were based on his review of a number of documents and exhibits, including well data for the Stork # 1 Well and the post-accident reports of appellees’ witnesses. Furthermore, Davenport provided his reasoning for concluding a blowout occurred in multiple formations within the Austin Chalk when he stated his definition for a formation. We [325]*325find that Davenport’s first two assertions are not simply conelusory.
Davenport’s third assertion is that his definition of a formation is the “most easily understood and generally accepted meaning of the term ‘formation.’ ” We find that Davenport’s industry definition is implicitly predicated upon his extensive work and educational experience in the oil and gas industry as a Ph.D. petroleum engineer — a list of accomplishments that Davenport incorporated into his affidavit in support of his reasoning. Though appellees provided deposition testimony from their own expert contradicting Davenport’s definition, they did not disqualify Davenport as an oil and gas expert, nor did they impeach him by deposing him directly and exploring the bases for his proposed definition. We find Davenport’s third assertion is not conelu-sory.
Appellees suggest that Davenport’s definition of formation is novel, created for this litigation, and should not be admitted to contravene a neutral and established dictionary definition. The proper means for attacking the reliability and relevance of expert testimony, however, is a Daubert challenge under Texas Rule of Civil Procedure 702. See generally Gam-mill v. Jack Williams Chevrolet, Inc., 972 S.W.2d 713, 726 (Tex.1998) (observing that the general principles of Rule 702, that expert testimony must be both rehable and relevant, apply to all expert testimony, including “opinions based on technical or specialized knowledge”). Although the trial court bears the ultimate burden of ensuring an expert’s testimony is rehable and relevant, “[t]he trial court’s gatekeeping function under Rule 702 does not supplant cross-examination as ‘the traditional and appropriate means of attacking shaky but admissible evidence.’ ” Id. at 728 (quoting from Daubert v. Merrell Dow Pharms., Inc., 509 U.S. 579, 596, 113 S.Ct. 2786, 125 L.Ed.2d 469 (1993))-. Here, however, ap-pellees did not raise a Daubert challenge in the trial court or on appeal, nor does the record reflect that they deposed Davenport to explore the foundations for his testimony. Therefore, the reliability of Davenport’s testimony under Rule 702 is not before us. In any case, although Davenport did not cite independent industry sources for his definition, his affidavit was expressly based on his extensive education and experience in the oil and gas industry. As the Gammill Court observed, “[experience alone may provide a sufficient basis for an expert’s testimony in some cases.” Gammill, 972 S.W.2d at 726.
We find that Davenport’s assertion that a “formation” should be interpreted as he defines it in his affidavit,' supported by his unimpeached qualifications, is sufficient to defeat summary judgment and establish a reasonable definition of a “formation” for this case. We conclude the policy is ambiguous.
Because we have found a formation under this policy is subject to at least two reasonable interpretations, one of which would render the March 23, 1994 event a coverable underground blowout within the meaning of the policy, we hold the trial court erred in granting appellees’ motion for summary judgment.
We sustain Mesealero’s point of error.
We reverse the summary judgment and remand the cause to the trial court.
Justice TAFT, dissenting.