Edward E. Simmons is an inventor who devised a strain sensitive element which is used in the measurement of impacts. He entered into a contract with California Institute of Technology concerning the payment of royalties under any licensing agreement made by him for its use and, subsequently, a contract with Baldwin Locomotive Works for the manufacture of his apparatus. The present action for *268declaratory relief was brought to have the first agreement declared rescinded, the second one declared void insofar as it confeis benefits upon the Institute, and for damages to the extent of benefits already received by the Institute. The appeal is from a judgment in his favor.
Following attendance at the Institute as a student, Simmons was graduated in 1934 and received his Master’s degree in 1936. During the summer and fall of the latter year, he occupied an office in the electrical engineering department on the campus and did work for various persons at the Institute and some others who had no connection with the organization. He was paid for his services upon an hourly basis.
At that time Dr. Donald S. Clark, a member of the faculty of the Institute, was in charge of a commercially sponsored project carried on by it and known as “Impact Research.” In the course of developing a method for measuring and determining the force of time relations which occur to metals during impact loading, Simmons, who had a reputation- for ingenuity, was consulted and made a suggestion for the use of a strain sensitive element. The evidence shows that Simmons’ ideas were used and, under his supervision, an apparatus embodying them was constructed which operated successfully.
A few months later, Simmons became a member of the staff of the Institute, on a part-time basis, for work with Impact Research but, at the time he conceived the idea of his invention and put it to practical use, he was not employed by, nor did he receive any compensation for services rendered to the research project. However, in June, 1938, Simmons was , employed by the Institute to work with Impact Research on a fellowship basis and received a reappointment for the following academic year.
In the fall of 1939, Baldwin Locomotive Works became interested in the Simmons invention and its representative had conversations with Simmons and Dr. Clark which culminated in the two agreements which are the basis of the present controversy. By this time, Simmons had become greatly interested in the work of Impact Research and its continuation, and suggested to Dr. Clark that the Baldwin offer would be a good proposition if the income received from royalties could be used to pay for materials and the expense of carrying on further research. Dr. Clark agreed that this would be most desirable.
Simmons and Dr. Clark had many conversations in regard *269to the terms and conditions which should be embodied in a contract and the provisions as to the use of the royalties. Dr. Clark told him that all royalties should be paid directly to the Institute but that they would be used for Impact Research. He explained to Simmons that the project could receive money only from the Institute, but that any sums paid by Baldwin would be used for Impact Research.
A memorandum entitled “Agreement,” bearing date of February 21, 1940, prepared by counsel for Baldwin and executed only by Simmons, recites “that as to my inventions involving an electrical strain gauge ... I will not grant any licenses to make, use or sell apparatus or methods embodying any of said inventions unless such licenses are approved by the California Institute of Technology, . . . any license so granted to be subject to such terms and conditions as may be reasonably required by [it] . . . The royalty on said inventions shall be paid to the . . . Institute . . . unless at my option I shall request in writing ... at least thirty (30) days prior to the end of any three (3) months’ period in which royalty accrues that up to and including forty per cent (40%) of the total royalty paid by a licensee shall be paid to me by the said licensee.' ’ Dr. Clark signed this document as a witness to the signature of Simmons, and it was received in evidence marked Exhibit “A.”
A contract executed by Simmons and Baldwin “as of March 13, 1940” declares that it is “subject to approval” of the Institute. This contract, designated in the record as Exhibit “B,” grants to Baldwin, with specified reservations, the exclusive license to make for use, rental and sale the inventions covered by the patent for a royalty of five (5) per cent of the sale price or income from rentals of that portion of apparatus embodying the invention. All royalty “shall be paid to the Institute, unless Simmons at his option shall request in writing to Baldwin and to the Institute, at least thirty (30) days prior to the end of any three (3) months’ period in which royalty accrues that any portion up to and including forty per cent (40%) of said royalty shall be paid to him by Baldwin.” Following this contract is a statement which reads as follows: “The granting of the foregoing license is approved by the California Institute of Technology in accordance with an agreement between [it and] Edward E. Simmons, Jr., . . . dated February 21, 1940. California Institute of Technology By: A. C. Baleh Pres.” It appears from the evidence that this approval was given by direction *270of the executive council of the Institute, but there was no action by it authorizing or ratifying a similar approval of the memorandum executed by Simmons.
Simmons testified that Dr. Clark had told him “the thing to do is to use the income from this invention of yours for the further development of impact testing and dynamic test procedures, and that . . . would make the continuation of the Impact Research project a possibility.” Simmons stated that his reply was: “I told Doctor Clark I was very much interested in having the Impact Research project continue, and I would be glad to have the income from this patent go to the Impact Research fund for further use in the development of equipment and tests in the laboratory. ’ ’ Dr. Clark assured Simmons that the royalties, although paid to the Institute would be used solely for the maintenance of Impact Research, and it appears from the evidence that this promise of continuing development of the project, made by Dr. Clark to the young scientist, was the major inducing force behind the agreements here in controversy.
Not until April or May of 1941 did Simmons know that Baldwin’s royalty payments were not being, and never had been, used in Impact Research. Simmons then called Dr. Clark’s attention to his promises on the subject. Shortly thereafter, Simmons was informed by Dr. Clark that his fellowship would not be renewed after July, 1941, because the project’s funds were running low. When Simmons pointed out that the sums being paid by Baldwin were more than the amount of his salary, Dr. Clark replied that the Institute, and not Impact Research, was receiving these royalties and he could do nothing about it. Moreover, said Dr. Clark, there was no need for any further discussion of the subject.
Although the project continued for some time after his discharge, Simmons was not again employed by it although subsequently, upon Dr. Clark’s recommendation, he did some work in connection with research being done by the Institute for the government. In January, 1942, Simmons served notice of rescission upon the Institute and this action followed.
Prom this and other evidence, the trial court determined that Simmons is entitled to the amount of the royalties received by the Institute, and it decreed that the Institute has no right to any royalties accrued, or to accrue. "This conclusion is based upon findings that Simmons received no consideration for the execution of the document designated Exhibit “A” nor for the licensing agreement (Exhibit “B”) *271insofar as the latter contract purports to confer rights upon the Institute; that Dr. Clark made fraudulent statements which Simmons believed and relied upon and that Dr. Clark’s acts and representations were neither authorized nor’ ratified by the Institute.
As justifying a reversal of the judgment, the Institute and Dr. Clark contend that there is insufficient evidence to support the findings as to lack of consideration for the benefits conferred upon the Institute, or the findings that the execution of the writings was induced by fraud. They also assert that no evidence should have been admitted upon the charge of fraud, and that Simmons is not entitled to the relief sought.
The finding as to the first cause of action is that “there was and is no consideration for the said written instrument executed by plaintiff and defendant Clark . . . [Exhibit A] ; that it is true that plaintiff has received nothing in consideration of, or for, said written instrument,” and that “it is true that . . . plaintiff executed and entered into the licensing agreement . . . referred to ... as Exhibit B, in pursuance and consideration of the aforementioned written instrument . . . Exhibit A; that it is true that there was, and is, no consideration for said licensing agreement, Exhibit B, insofar as said agreement purports to confer rights upon California Institute of Technology; that it is true that plaintiff has received nothing in consideration of, or for, said licensing agreement insofar as it purports to confer rights upon . . . California Institute of Technology.” Another finding was stated in the following terms: “That it is not true that the aforementioned instrument . . . Exhibit A, was at any time and/or now is, supported by a valid and/or valuable consideration, or any consideration whatsoever; that it is not true that the aforementioned licensing agreement . . . Exhibit B, insofar as the same purports to confer rights upon . . . [the] Institute, was at any time and/or now is, supported by a valid and/or valuable consideration, or any consideration whatsoever.”
The findings clearly show that the trial court treated Exhibit “A” as a memorandum of a contract between Simmons and the Institute resulting from the negotiations between him and Dr. Clark. Exhibit “B” is a third party beneficiary contract between Simmons and Baldwin, under which the Institute is the beneficiary.
The finding that Exhibit “A” was executed by Sim*272mons without consideration on the part of the Institute is supported by the evidence. The true consideration of a contract may be shown by extrinsic evidence (Shiver v. Liberty Bldg.-Loan Assn., 16 Cal.2d 296 [106 P.2d 4]; Johnston v. Courtial, 216 Cal. 506 [14 P.2d 771]), and the record includes testimony to the effect that both Simmons and Dr. Clark understood the phrase, “in consideration of employment,” which appears in Exhibit “A” as referring to past employment. Such employment is inadequate consideration to support a contract, and the promises of Simmons to grant licenses to use his invention only upon approval by the Institute and to pay royalties to the Institute were made without any counterpromise by the Institute.
The appellants argue that there are three “aspects of consideration,” any one of which is sufficient to support the promises of Simmons. First, it is said that Simmons was reemployed by the Institute after the term under which he was serving had expired. Secondly, the appellants assert that the agreements were made as a compromise of a dispute between Simmons and the Institute regarding the Institute’s claim to an interest in the patent. Finally, they rely upon the services rendered by the Institute, through Dr. Clark, in handling the negotiations of the contract with Baldwin as constituting sufficient and adequate consideration.
But the consideration for a promise must be an act or a return promise, bargained for and given in exchange for the promise. (Bard v. Kent, 19 Cal.2d 449 [122 P.2d 8, 139 A.L.R 1032]; Tiffany & Co. v. Spreckels, 202 Cal. 778 [262 P. 742]; Williams v. Hasshagen, 166 Cal. 386 [137 P. 9]; Lasar v. Johnson, 125 Cal. 549 [58 P. 161]; Rest., Contracts, § 75; see Williston, Contracts, [rev. ed.] §§ 61,100,102, 102a.) In the words of section 75 of the Restatement of Contracts (com. b) : “Consideration must actually be bargained for as the exchange for the promise . . . The existence or non-existence of a bargain where something has been parted with by the promisee or received by the promisor depends upon the manifested intention of the parties . . . The fact that the promisee relies on the promise to his injury, or the promisor gains some advantage therefrom, does not establish consideration without the element of bargain or agreed exchange.” (Language approved in Bard v. Kent, supra, p. 452.) In the present ease, the evidence shows, and the trial court necessarily found, that it was the past employment of Simmons, and not the promises or acts suggested by the appellants, *273which the parties contemplated as the consideration for Simmons’ promises. His promises, therefore, were made gratuitously and may be revoked by cancellation or rescission of the instruments. (Baltimore & Ohio R. Co. v. Evans, 188 F. 8 [110 C.C.A. 156]; 1 Black on Rescission and Cancellation, (2d ed.) § 157, p. 458 et seq.)
The trial court found that, insofar as the third party beneficiary provision of Exhibit “B” is concerned, it was executed “in pursuance and consideration of” the contract evidenced by Exhibit “A.” It necessarily follows that if the prior contract was rescinded for lack of consideration, the beneficiary provisions of Exhibit “B,” which were based thereon, are unsupported by consideration and also may be rescinded. The judgment was rendered upon this theory.
The findings of the.trial court in favor of Simmons on the issue of fraud are likewise supported by the record. It clearly appears that, apart from the question of the consideration bargained for, Simmons was induced to enter into the contracts in reliance upon certain misrepresentations by Dr. Clark. As a young man, experimenting on the borderland of what appeared to be a vast new field of scientific and engineering development, he was greatly impressed with the importance of the work being conducted at Impact Research. “It was a new field,” he explained, “the matter of dynamic testing materials was just becoming of importance at that time, and I was extremely interested in that particular subject, and the instrumentation useful in such investigations, and I think we mutually agreed that the money could be well used in Impact work.”
The substantial evidence which supports these findings places them beyond the reach of an appellate court. (DeYoung v. DeYoung, 27 Cal.2d 521 [165 P.2d 457]; Viner v. Untrecht, 26 Cal.2d 261 [158 P.2d 3]; Juchert v. California Water Service Co., 16 Cal.2d 500 [106 P.2d 886]; Raggio v. Mallory, 10 Cal.2d 723 [76 P.2d 660]; Albaugh v. Mt. Shasta Power Corp., 9 Cal.2d 751 [73 P.2d 217]; Gane v. Gane, 6 Cal.2d 145 [56 P.2d 948]; Bellon v. Silver Gate Theatres, Inc., 4 Cal.2d 1 [47 P.2d 462]; Crawford v. Southern Pac. Co., 3 Cal.2d 427 [45 P.2d 183].) Moreover, the determination that Dr. Clark was employed by the Institute, but that his acts and representations were not authorized or ratified by it is not inconsistent with other conclusions in favor of Simmons, for “the rule is now firmly established that a rescission of con*274tract may be had against even an innocent principal because of unauthorized misrepresentations by his agent.” (Greenberg v. DuBain Realty Corp., 2 Cal.2d 628, 629 [42 P.2d 628]; and see Weiner v. Roof, 10 Cal.2d 450, 453 [74 P.2d 736].)
The Institute and Dr. Clark assert that it was error to have admitted parol evidence of such promissory fraud, because it is “at variance with the contractual features of the writing.” The determinative question upon this issue is “whether the writing was intended to cover a certain subject of negotiation; for if it was not, then the writing does not embody the transaction on that subject, and one of the circumstances of decision will be whether the one subject is so associated with the other that they are in effect ‘parts’ of the same transaction, and therefore, if reduced to writing at all, they must be governed by the same writing.” (9 Wig-more on Evidence, § 2430, p. 97.)
It is clear that “where the execution of a contract has been induced by a promise made without any intention of performing it, this constitutes such fraud in obtaining the contract that it may be declared null and void” but that “a distinction must be made between ... a parol promise . . ., which by its very nature is superseded by the final writing, inconsistent with it, and a promise made with no intention of performing the same, not inconsistent with the writing, but which was the inducing cause thereof.” (Cobbs v. Cobbs, 53 Cal.App.2d 780, 783, 785 [128 P.2d 373].) Exhibit “A” provided that “the royalty on said invention shall be paid to the . . . Institute, or committees or other agencies or bodies which the . . . Institute may designate. ...” The promises of Dr. Clark were that the money, although paid to the Institute as a matter of administrative procedure, nevertheless was to be used exclusively for Impact Research. Therefore, the promise was directed to the matter of the use of the money, whereas the terms of the memorandum dealt with nothing more than the form of the payment of it. These promises by Dr. Clark as to the use of the royalties were the fraudulent inducement, or motive, for the contract, but they were not incorporated in or superseded by the terms of the agreement as to payment. The two are not inconsistent or “at variance,” inasmuch as they deal with wholly different matters. It was, therefore, proper to receive parol evidence to prove the promises of Dr. Clark. (Stockburger v. Dolan, 14 Cal.2d 313 [94 P.2d 33, 128 A.L.R. 83]; Buckner v. Leon & Co., 204 Cal. 225 [267 P. 693]; Whittier v. Home Sav. Bank, *275161 Cal. 311 [119 P. 92]; Savings Bank of So. Cal. v. Asbury, 117 Cal. 96 [48 P. 1081]; Lindsay v. Mack, 5 Cal.App.2d 491 [43 P.2d 350]; Gardiner v. Burket, 3 Cal.App.2d 666 [40 P.2d 279].)
The argument of the appellants that Simmons is not entitled to the relief which he seeks is in two parts. The first, directed'to Exhibit “A,” is that he failed to tender or return the benefits received. This argument assumes that something other than past employment was the intended consideration for the promises of Simmons. But as he received no benefits in return for his promise, he is not required to tender or return any amounts paid to him. The second part of the argument, directed to Exhibit “B,” is that a rescission must be total and not partial; that Simmons may not rescind as to the burdens imposed upon him while retaining the benefits.
The general rule is that one must rescind all of his contract and may not retain rights under it which he deems desirable to have and repudiate the remainder of its provisions. (Buena Vista Fruit Co. v. Tuohy, 107 Cal. 243 [40 P. 386]; Bohall v. Diller, 41 Cal. 532; Purdy v. Bullard, 41 Cal. 444; Jones v. Bay Cities Electric Co., 22 Cal.App. 81 [133 P. 492].) The theory underlying such a rule is that retention of only the benefits of the transaction amounts to unjust enrichment and binds the parties to a contract which they did not contemplate.
This rule, however, is not controlling in the case of a severable or divisible contract such as Exhibit “B.” Generally speaking, the test of whether a contract is divisible is that if the consideration is single, the contract is entire, but if the consideration is apportioned, the contract may be regarded as severable. (Traiman v. Rappaport, 41 F.2d 336, 71 A.L.R. 475.) And a contract may be severable as to some of its terms, or for certain purposes, but indivisible as to other terms, or for other purposes. (In re Marshall’s Garage, 63 F.2d 759; Ireland v. Craggs, 56 F.2d 785; Cosden Oil Co. v. Scarborough, 55 F.2d 634.) Here there was a total rescission of the provisions relating to the Institute, and they are clearly severable from the terms and conditions applicable to Baldwin. And “where a good cause for rescission exists as to one part of a divisible or separable contract, such portion may be rescinded in equity without disturbing the remainder.” (Hesselberg v. Aetna Life Ins. Co., 102 F.2d 23, 25.) The rescission as to the Institute is total, not *276partial. No injustice is done the Institute by Simmons’ failure to rescind as to Baldwin, and the Institute has incurred no injury if Simmons and Baldwin, as between themselves, elect to be bound by Exhibit “B.” By the terms of the decree of the trial court, the Institute and Simmons have been placed in status quo.
As to royalties already received by the Institute, they were paid under a contract induced by fraud and executed in the mistaken belief that, under the terms of Exhibit “A,” Simmons owed an obligation to the Institute. In such a situation the principles of unjust enrichment and restitution require that such monies may not be retained. (See Restatement of Restitutions, §§ 5, 133[1].)
There is no merit in the argument, made for the first time upon appeal, that the judgment is improper because Baldwin was not made a party to the action. Exhibit “A” is a memorandum of an agreement between Simmons and the Institute; Baldwin has no rights or interests dependent upon or arising from that agreement. It is, therefore, only as the judgment may purport to affect Exhibit “B,” the contract between Simmons and Baldwin, that the licensee could have any direct interest. However, the trial court made no finding as to Baldwin’s interests under that contract; it expressly restricted the determination of the rights as between Simmons and the Institute to the amount of royalties paid or to be paid pursuant to the contract. Baldwin continues to receive the benefit it contracted for, the right to manufacture the invention; it is under obligation to pay the same amount in royalties for that privilege and probably it is of no consequence to Baldwin to whom that money is paid. If it is, that company may raise such defense in any action hereafter brought by Simmons to collect royalties.
The interests of Baldwin under the contract are “so separable that a decree may be rendered between the parties before the court without affecting . . . [them, so that they] may perhaps be ‘necessary’ parties to a complete settlement of the entire controversy or transaction, but are not ‘indispensable’ to any valid judgment . . . They should normally be joined . . . [b]ut, since the rule ... [as to joinder of necessary parties] is one of equity, it is limited and qualified by considerations of fairness, convenience, and practicality. Where, for example, it is impossible to find these other persons or impracticable to bring them in, the action may proceed as to those parties who are present.” (Bank of Cali*277fornia v. Superior Court, 16 Cal.2d 516, 523 [106 P.2d 879].)
The fact that Baldwin is a foreign corporation may have placed some difficulty in the way of making it a party to this litigation had Simmons been disposed to do so. But he was satisfied to proceed solely against the Institute and Clark and must accept the responsibility of a judgment which does not bind Baldwin. Any dispute between Simmons and Baldwin has not been settled by this judgment, and Baldwin is in no way bound by it.
The judgment is affirmed.
Gibson, C. J., Traynor, J., and Peters, J. pro tern., concurred.