The opinion of the court was delivered by
This is an appeal from the judgment of the district court of Dickinson County, Kansas, finding the joint will of Herbert Dix Chronister and his wife, Mabel Belle LaForce Chronister to be contractual. These parties will hereafter be referred to as Herbert and Mabel.
The facts are essentially undisputed. On September 4, 1951, Herbert and Mabel Chronister executed a joint will in which they bequeathed all property owned by them either jointly or severally to the survivor of them, for his or her own personal use and benefit forever, and after the death of the survivor directed their executors to distribute the remaining household furnishings and effects to five beneficiaries, and to sell the remainder of their property, both real and personal, at public or private sale, and distribute the proceeds thereof to the five named beneficiaries, who are shown by the record to be nieces and nephews of Herbert. We shall refer to the will in more detail in the progress of this opinion.
Herbert died September 16, 1951, and on Mabels petition the joint will was admitted to probate by the Butler County Probate Court as Herbert’s will. Herbert’s estate was settled November 6, 1952, at which time the probate court found that Mabel was entitled to the estate during her lifetime with power of disposal, and assigned all property, real and personal, to Mabel to be “held and used or disposed of during her life time, and any estate remaining at the time of her death shall upon proper proceedings be assigned to [the five beneficiaries].”
On August 24, 1965, Mabel executed a second will after having first consulted with John Lehman, an Abilene attorney. In this will Mabel bequeathed 562 shares of Colgate-Palmolive Company stock valued at some $13,000 to her sister, Edith Kessinger, and her *368children, share and share alike, and devised and bequeathed the rest and residue of her estate to the same nieces and nephews of Herbert named in the joint, or first, will. Mabel died August 22, 1966, and her will was offered for probate in the Dickinson County-Probate Court two days later. As a defense to the admission of Mabel’s will to probate, the executors named in the joint will filed an answer in which they prayed that Herbert and Mabel’s joint will be allowed as Mabel’s last will and testament, and that it be enforced against her estate as a contract.
The cause was thereafter transferred to the district court where, on motion of the executor of Mabel’s will, the district court entered summary judgment admitting her will to probate but reserving jurisdiction to determine whether the joint will of Herbert and Mabel was contractual. At a subsequent hearing on the contractual issue alone, the trial court found the joint will to be contractual on its face and entered judgment accordingly. Mabel’s sister, Mrs. Kessinger, and her two children have appealed from this judgment.
Ry way of preliminary comment, we pause here to reiterate what was said in Menke v. Duwe et al., 117 Kan. 207, 216, 230 Pac. 1065, that a single instrument may be both a will contractual in character, and a contract testamentary in nature; as a will it is revocable but as a contract it is enforceable; and although a contractual will revoked by execution of a second will, cannot be probated, it may nonetheless be enforced as a contract against the estate of the testator breaching it.
The respective positions of the parties to this appeal may be summarized, essentially, in this way: the appellees, who will be referred to herein collectively as the Chronisters, say the will is not ambiguous, that it is contractual on its face and that extrinsic evidence is not admissible to refute its contractual character; while the appellants, to whom we shall hereafter refer as the Kessingers, contend the will is ambiguous, that it does not show up on its face that it is based on contract, and that extrinsic evidence is admissible to establish there was no agreement or contract.
There has been no dearth of litigation in this state over wills which were claimed to be contractual. At first glance, some of our pronouncements on the subject may appear to follow somewhat divergent paths, but more assiduous study should reveal to a discerning eye several guide posts placed along the route to direct a wary traveler to and along the true highway. While we do not *369propose to discuss each of our former decisions relating to contractual wills, we think several must be noted as having special significance in the emergence of principles which are controlling in this area.
It may first be relevant to point out that disputes may arise as to whether or not a will is contractual, not only when the will is joint in form and contained in a single common instrument, but also when it is drawn in the form of a separate document. No matter which class it comes under, a will may be claimed to be the product of a pre-existing agreement. (See Frontier Lodge v. Wilson, 139 Kan. 75, 30 P. 2d 307; In re Estate of Pennington, 158 Kan. 495, 148 P. 2d 516; Eikmeier v. Eikmeier, 174 Kan. 71, 254 P. 2d 236.) Unless otherwise noted, cases cited throughout this opinion fall within the first category.
In a comparatively early case, Lewis v. Lewis, 104 Kan. 269, 178 Pac. 421, a will which had been executed jointly by a husband and wife was held by this court to be contractual on its face. In response to a contention that there was no evidence to establish that the testators had entered into any contract to make a will, the court replied:
“. . . How could such a will be voluntarily executed if there was no agreement or understanding that it would be made? The will itself, its terms, and its execution, are evidence that such a contract was made. . . (p. 273.)
Since the Lewis case was decided, this court, on a number of occasions, has construed a will which was jointly executed as establishing, per se, a contractual character. In a few instances, specific reference to an existing agreement was set out in the will itself. (Warwick v. Zimmerman, 126 Kan. 619, 270 Pac. 612; Berry v. Berry, 168 Kan. 253, 212 P. 2d 283; In re Estate of Buckner, 186 Kan. 176, 348 P. 2d 818.) In other cases, where specific reference to a prior contract was lacking, wills have been construed as contractual on the basis of specific provisions or terms, one of the common denominators usually being a provision for the disposition of property after the death of the survivor. See, In re Estate of Adkins, 161 Kan. 239, 167 P. 2d 618; Beall v. Hardie, 177 Kan. 353, 279 P. 2d 276; In re Estate of Weidman, 181 Kan. 718, 314 P. 2d 327.) Such a provision was found among the provisions of the will in the Lewis case.
Three opinions of recent origin deserve particular mention, for within their boundaries are mirrored the essence of our decisions in *370the domain of joint contractual wills. In the case of In re Estate of Miller, 186 Kan. 87, 348 P. 2d 1033, this court, speaking through Mr. Justice Schroeder, reviewed many of our earlier decisions and said:
“The existence or nonexistence of an agreement or contract is in its very nature a fact. In the cases where wills have been held to be contractual, it will be found that either the wills expressly recited that they were contractual, or the fact that they were contractual was established by evidence. (In re Estate of Pennington, 158 Kan. 495, 148 P. 2d 516; and Frontier Lodge v. Wilson, 139 Kan. 75, 30 P. 2d 307. . . . (p. 94.)
“Reciprocal provisions in mutual wills are insufficient to make the will contractual. The general rule is that the execution of a joint will is not of itself sufficient evidence of an enforceable contract to devise between the testators, so as to make the contract enforceable in equity. . . .” (pp. 95, 96.)
In discussing Lewis v. Lewis, supra, Justice Schroeder, after first observing that the language used in that case was rather broad, and that the court had made specific reference in its opinion to “the will itself, its terms,” proceeded to say:
“The case, in effect, holds that where a joint and mutual will is executed by a husband and wife the will itself and its terms may be taken into consideration as circumstantial evidence upon which to base a finding that the will is contractual. This is not to say that the execution of a joint and mutual will compels such an inference.” (p. 97.)
The conclusion of this court in the Miller case appears on page 98 of the opinion:
“A reading of the joint and mutual will of James and Elsie in the instant case shows it to be wholly lacking in any language which would indicate that it was contractual. If it was the intention of the parties to have a contract, it would have been an easy matter to have worded the will to express such an intention. How can it be said that James and Elsie, [the testators] contracted to leave their property to Lucille [the asserted remainderman] upon the death of the survivor when the will itself contains no provision to cover this eventuality? . . .” (Emphasis supplied.)
It is to be noted that extrinsic evidence had not been offered in the Miller case as to whether the will was contractual, and the existence or nonexistence of an antecedent agreement had to be determined from the four corners of the will itself. Consequently, this court was constrained to hold that no agreement was established.
In the second case, In re Estate of Tompkins, 195 Kan. 467, 407 P. 2d 545, a husband and wife executed a joint will leaving all their property to the survivor, with right of disposal, and providing for disposition thereof upon death of the survivor. A dispute having arisen over whether the will was contractual, the trial court held *371that the terms of the will itself showed it to be joint, mutual and contractual. On appeal this court affirmed the judgment of the trial court in these words:
“. . . A careful examination of the language used can only lead to the conclusion that the will in this case was the result of an agreement or contract of the testators. . . .” (p. 471.)
In arriving at this conclusion we fingered these factors: the use of the plural words “we” and “ours” throughout the will; the directions contained for the disposition of the testators’ property after the death of the survivor; and the carefully drawn provisions for the disposition of any share in case of a lapsed residuary bequest.
The admissibility of extrinsic evidence to show the contractual nature of a will was also raised in Tompkins, where the scrivener of the will had testified without objection concerning conversations had over the years with the joint testators about how their estate was to be planned. As to this testimony we said:
“In this connection it is to be noted that this court has often stated that the rule that parol evidence is never admissible to change or vary the terms of an unambiguous will does not render inadmissible extrinsic evidence that a will was executed pursuant to an agreement. . . .” (p. 474.)
However, as to additional testimony given by the scrivener and admitted over objections, regarding his opinion of the contractual nature of the will, and his explanation given to the testators when the will was executed, we specifically pointed out:
“. . . In so far as such testimony may be said to tend to change or alter the provisions or terms of the will it was inadmissible. (Guthrie v. Guthrie, 130 Kan. 433, 286 Pac. 195; In re Estate of Schnack, 155 Kan. 861, 130 P. 2d 591; In re Estate of Sowder, 185 Kan. 74, 340 P. 2d 907.) . . .” (p. 474.)
The third and latest case is In re Estate of Wade, 202 Kan. 380, 449 P. 2d 488. Here, extrinsic evidence was introduced at the hearing before the trial court and the court made findings of fact therefrom which we held compelled a conclusion that the will was of a contractual character. It is clear from a reading of the Wade will that it did not possess the plain and unambiguous qualities of the will being considered in Tompkins. On page 390 of the opinion, Mr. Justice Kaul, speaking for the court declared:
“In the Wade will there is no language stating it to be contractual, nor does the language used supply with clarity the evidence of an underlying obligation. . . .”
This being true, reliance had to be placed on extrinsic evidence in determining the intent of the testators and the contractual or *372non-contractual character of the joint will. We stressed again, as we did in Miller, that whether a will is the product of an agreement is a question of fact, which must be established by proof; and that sufficient proof is not supplied by the fact, alone, that a will is joint and mutual, even though a will itself and its terms may suffice as circumstantial evidence to form the basis for finding that a will is contractual.
We distinctly pointed out in Wade that the Tompkins will was held to be contractual on its face because its language clearly indicated the elements of an undertaking by which the testators agreed to be bound. “Although extrinsic evidence tending to prove a contract was received in Tompkins,” we said, on page 389, “it was described as unnecessary for the reason the will was contractual on its face.”
From the cases cited, we believe these rules may be deduced: (1) Whether a will is contractual, be it a joint will or one of separate wills, is a question of fact which must be established by proof. (2) The mere fact that a will is joint does not in and of itself establish it to be the result of a pre-existing agreement. (3) A joint and mutual will and the terms and provisions thereof, may be considered sufficient as circumstantial evidence to establish that it was executed pursuant to an agreement. (4) Where a joint will shows on its face by the terms and provisions thereof that it is contractual in character, extrinsic evidence is not admissible for the purpose of proving otherwise. (5) Where there is ambiguity from the language used in a joint will as to whether or not it is based on a contract, extrinsic evidence is admissible to establish either the existence or nonexistence of a contract.
Which brings us to this crucial point: Did the trial court err in construing the joint will of Herbert and Mabel Chronister as being contractual on its face? Although portions of the will have already been mentioned, we are impelled to summarize all pertinent provisions, at this point, even though some repetition will inevitably result.
In clause two, the testators devised and bequeathed to the survivor of them all property jointly or severally owned by them for the survivor’s personal use and benefit forever. Clause three directed distribution of household effects to the five Chronisters, in the executors’ absolute discretion, and directed the sale of all other property. Clause four directed distribution of the sale proceeds *373equally to the Chronisters, and further provided that should any beneficiary die prior to the death of the surviving testator, his or her share should be distributed to his or her issue, but should no issue survive, then such share should go to the remaining beneficiaries, or their surviving issue. Clause five provides that those friends or heirs not mentioned in the will were omitted with full knowledge thereof.
In its Memorandum Decision finding that the joint will was contractual, the trial court included the following pregnant paragraphs:
“11. The court has used the rule of construction that the intent of the testator is paramount and that all other rules are subordinate thereto. (In re Estate of Tompkins, 195 Kansas 467, page 471.)
“12. The will dated 4 September, 1951, especially the following language used therein, was considered by the court.
“(a) The use of the plural pronouns in the declaration (i.e., ‘we’, ‘our’, and ‘us’).
“(b) The use of plural ponouns [sic], without exception, in the body of the will where said will speaks of a bequest or devise.
“(c) Identical disposition of any share in case of a lapse of any residuary bequest resulting from the death of any beneficiary.
“(d) A mutual exclusion (paragraph 5) of heirs of either of the parties or their friends not mentioned therein.
“(e) The use of the word ‘forever’ in the devise and bequest to the survivor in paragraph 2.
“13. The court believes that the term ‘forever’, if standing alone, as used in the survivor paragraph would denote a fee devise, but when the will is construed in toto the court can reach no other conclusion than that there is a clear indication of an agreement by the testators as to the terms and provisions of the will, and that they both intended to be bound thereby.”
To the factors pinpointed by the trial court as compelling the conclusion that the will is contractual, we would add only the fact, which we deem of much significance, that full and explicit provision is made for disposition of the testators’ remaining property after the death of the survivor. As we have said before, provisions of like nature have commonly been said to evidence, in the case of joint wills, an understanding between the joint testators by which they intended to bind themselves.
In agreeing with the trial court’s judgment that the joint will is not ambiguous, but shows by its own terms and provisions a contractual background, we wish to make clear that we refer to lack of ambiguity only in relation to the issue of whether the will was executed pursuant to agreement. Yet the rules with respect to determining whether a testamentary document is ambiguous or not *374ambiguous as to its contractual character can be no different, in our judgment, than when the document is tested for ambiguity, or lack thereof, as to the meaning of its provisions disposing of property. In either case the critical test is whether the intention of the testator, or testators as the case may be, can be gathered from the four corners of the instrument itself. If the testamentary intention can be gathered from the face of the will, ambiguity does not exist; otherwise it does. In the Tompkins case, we phrased the rule in this fashion:
“The firmly established rule in this jurisdiction for the construction of wills, to which all other rules are subordinate, is that the intention of the testator as garnered from all parts of the will is to be given effect, and that doubtful or inaccurate expressions in the will shall not override the obvious intention of tifie testator. . . .” (p. 471.)
The will of Herbert and Mabel Chronister being contractual, as shown by the terms and provisions appearing upon its facé, it follows that extrinsic evidence was not admissible to disprove such fact, for that issue was foreclosed. Accordingly, the trial court did not err in excluding the proffered extrinsic evidence. It is familiar and respected law that parol or extrinsic evidence is not competent to explain or to vary the term of a written instrument which is clear and unambiguous upon its face. (Baldwin v. Hambleton, 196 Kan. 353, 411 P. 2d 626; In re Estate of Smith, 199 Kan. 89, 427 P. 2d 443; In re Estate of Tompkins, supra.)
Considerable emphasis is placed by the Kessingers (appellants) on Menke v. Duwe et al., supra, where, after considering extrinsic evidence relating to the circumstances surrounding the execution of a joint will, including conversations occurring at that time, the trial court found the will was not executed pursuant to an agreement. This court, on appeal, affirmed that finding but the decision must be read in the light of the form in which the will was prepared and the terms it contained. The will differs markedly in form from the Chronister will, and to a somewhat lesser degree in content. Although but one instrument was executed by the testators, its arrangement as to form, as well as its wording, strongly indicated separate wills contained within a single document. On page 220 the court sagely observed that the instrument itself did not compel the inference that it was contractual in character.
The Kessingers further contend that the probate court, in its order of final settlement of Herbert’s estate, construed the joint will *375as non-contractual, and they say this construction is res judicata inasmuch as no appeal was taken from the order. Their ingenious argument goes like this: During probate, the executor of Herbert’s estate delivered to Mabel, under court order, certain stocks held by Herbert and Mabel in joint tenancy; in its order of final settlement the probate court approved all acts of the executor; in Berry v. Berry, supra, this court held the execution of a joint contractual will terminates a joint tenancy; ergo, the probate court, in approving the delivery of the stocks to Mabel, construed the will as not being contractual, for otherwise, say the Kessingers, the stocks would have been retained in Herbert’s estate.
We decline to follow this somewhat tortured reasoning. Particularly are we loath to declare judicially that the Butler County Probate Court construed the will as- non-contractual when the court actually found Mabel entitled to the estate during her lifetime with power of disposal, and assigned the same to her for life with power of disposal and assigned the estate remaining to the Chronisters. Nor are we convinced that a finding by the probate court that the will was not contractual, had it been made, would be binding on the Chronisters. In 169 A. L. R. 82, we find this statement:
. . The probate of an instrument executed jointly by two persons, upon the death of one of them, as his will, is not an adjudication that the will is not contractual or that the survivor is not bound by its provisions.
“. . . An order closing the estate of a testate decedent and approving the distribution of the residue of his estate pursuant to the terms of his will to •one surviving him is not an adjudication which precludes the assertion subsequently of a claim of the violation by the survivor of her agreement with the testator for the execution of separate wills containing certain specified provisions. . . .”
It has been suggested that the probate court’s order of final settlement was actually a res judicata pronouncement that the will was contractual in character. In view of the disposition we are making •of this case we do not feel called upon to determine this point.
We conclude that no error was committed by the trial court, and its judgment is affirmed.