writing is of a higher nature than matter of averment is, as we have seen, that where the two come into conflict the higher must prevail. Of this rule there are several important consequences, which, though familiar enough, will be shortly discussed for the purpose of exhibiting their relation to each other as illustrations of a single principle.

I. The first of these is the rule that a deed cannot be annulled or altered except by deed. This is clearly the result of the general principle of the superiority of matter in writing. If the plaintiff proves his claim by a deed, the defendant (if he admits the deed) must prove his defence by deed also; otherwise it would be averment against specialty, and the latter would by rule of law prevail. Thus in 20 Edward I1 a plaintiff says: "We have put forward a deed which is admitted in court, and you have nothing in hand to certify the court of the truth of your statement, but only make an assertion; judgment as of undefended.' Therefore if the defence is that the deed has been by subsequent agreement annulled or altered, the defendant must produce a deed in support of his statement. This rule of evidence that a deed cannot be proved to have been released or altered except by deed, has naturally passed over into the rule of substantive law that a deed cannot be released or altered except by deed. This rule is commonly expressed in a more or less modified form of Ulpian's maxim: Nihil tam naturale est quam eo genere quidque dissolvere quo colligatum est. Thus we read: Quomodo quid constituitur eodem modo dissolvitur; a Record by a Record, Writing by Writing, Parliament by Parliament, Parol by Parol3. This however is rather the adaptation of a formula than the adoption of a rule of Roman law. The maxim in question was applied systematically throughout the Roman law of contracts: a contract per aes et libram must be released per aes et libram, a stipulation by acceptilation, and consensual contracts by simple consent. There seems no evidence however as to whether the Roman rule originated in any principle of evidence, or was merely deduced from jus naturale by the ingenuity of Roman jurists. The English rule on the other hand is, in its most important applications, clearly deducible from the principle of the relative value of different kinds of evidence.

6

This rule has not of course been extended beyond deeds. It is well settled that at Common Law a simple contract, even though reduced to writing, may be rescinded or modified by parol.

II. The second result of the superiority of matter of specialty

1 Y. B. 20 & 21 Ed. I. 64; see also Y. B. 32 & 33 Ed. I. So & 136.

2 D. 50, 17, 35.

Jenk. 2 Cent. Case 40; see also Shep. Touch., by Preston, 323.

relates to the defences of payment and of accord and satisfaction. Mere payment was at Common Law no plea to an action on a single bond, for that would have been to allow matter of averment to prevail against matter of specialty. Since the plaintiff proved his debt by specialty, the defendant must prove by evidence of as high a nature that the debt was no longer owing. Accordingly it is said in 34 Edward I1: Since we have your deed, which you have admitted, and which binds you in that debt, and you to loose yourself from that debt have no acquittance but only an assertion, we pray judgment.' In 20 Edward I on the other hand there is an anomalous case in which a tally was allowed in spite of protest to be set up in proof of payment of a recognizance. John asked what Robert had to show for the payment. Robert said: See here a tally. John: We think that a recognizance made in the King's Court is the highest matter (la plus haute chose) transacted in Court; so we pray judgment if by a tally which can be defeated by good law, he can defeat the recognizance. Metingham J.: It may be that Robert had no clerk at hand; therefore answer if the tally be your deed or not. . . . John made his law twelve-handed, whereupon Robert was sent to prison for forging the tally 2. In Doctor and Student we are told that the rule in question is ordained by the law to avoid a great inconvenience that else might happen to come to many people; that is to say, that every man by a nude parol and a bare averment should avoid an obligation; wherefore to avoid that inconvenience the law hath ordained that as the defendant is charged by a sufficient writing, that so he must be discharged by sufficient. writing or by some other thing of as high authority as the obligation is. Notwithstanding however the approval of the Student of the Common Law, the rule that payment without acquittance could not be pleaded to a single bond was abolished by 4 & 5 Anne c. 3. It may be observed that the rule never applied to covenants, for there the cause of action is based not solely on a deed but also on matter in pais, namely the breach of the covenant, and therefore matter in pais is admissible in answer. Neither did the rule apply to the condition of a bond, the performance or non-performance of the condition lying it was said in pais. Sans acquitance home pledera payment de money paiable sur endorcement d'un obligation.'

III. The third application of the rule that matter of specialty must prevail over matter of averment is the doctrine of estoppel by deed.

1 Y. B. 33-35 Ed. I. 330; see also 20 & 21 Ed. I. 304; 21 & 22 Ed. I. 410; 32 & 33 Ed. I. 184.

2 Y. B. 20 & 21 El. I. 330.

Blake's Case, 6 Rep., 43 b.

3 Doctor and Student, I. 12.

Y. B. 12 Henry IV. 23. pl. 6; see Y. B. 21 & 22 Ed. I. 540.

Estoppel by record has, as we shall see, an analogous origin, though the modern equitable estoppel, or estoppel by conduct, is founded on a totally distinct principle. Speaking generally the rule of estoppel by deed is that matters of fact acknowledged in a deed to be true cannot be denied to be so by the maker of the deed or by those claiming through or under him. The structure that medieval lawyers constructed on this simple foundation is one of the many marvels of English law. This rule, like the others that we have discussed, is based on the conclusive presumption of law that verbal evidence is inferior to written. The conflict however between this presumption and actual fact has come out more strongly in the case of estoppel than in any of the other applications of the principle. It became indeed so evident that the rule of estoppel by deed frequently shut out the truth, that its effect in so doing actually came to be regarded as its characteristic element. Thus in the familiar words of Lord Coke: 'It is called an estoppel or conclusion because a man's owne act or acceptance stoppeth or closeth up his mouth to alleage or plead the truth'.' But the older estoppels were certainly meant to establish the truth, not to subvert it, and with respect to them such a definition is mere satire. The modern estoppel by conduct however has been established with the deliberate intention of making falsehood take the place of truth. It is curious that two principles established with exactly opposite intentions should fall within the same definition. The failure of the original theory on which estoppel by deed was based has led to attempts being made to discover some new and surer foundation for that rule. The modern tendency is to regard it as founded on contract, as arising in fact from a binding agreement to admit certain facts as true 2.

'Estoppel against estoppel puts the matter at large.' Where there is a conflict of evidence, the higher prevails by rule of law; but where the evidence on each side is in the eye of the law of equal rank, there the matter is at large, that is, is left to the discretion of those with whom the decision rests. Therefore specialty against parol raises an estoppel; the former necessarily prevails. But parol against parol or specialty against specialty leaves the matter at large.

Estoppel by deed is a doctrine established from the earliest times. In 22 Edward I we are told that nothing contained in a writing can by any exception of the parties be removed.' And again in the same year it is argued that the maker of a deed is estopped by its date: 'You cannot say that; for you have admitted the deed 2 See Horton v. Westminster Commissioners, 7 Ex. 791. Y. B. 21 & 22 Ed. I. 436.

1 Co. Litt., sect. 667.

and consequently whatever is contained in the deed'.

No maxim

is commoner in the early reports than that a man cannot contradict his own deed, but its application is more generally an illustration of the other branches of the general rule than of estoppel in the strict sense.

In certain miscellaneous instances estoppel has passed from a rule of evidence into a rule of substantive law. It is obvious that such a principle has a tendency to alter the substantive law. For if A is the legal consequence of B, and C is made conclusive evidence of B, there is a tendency for the element B to be lost sight of, and for C itself to be regarded as the legal antecedent of A. A bond is an undoubted instance of such a process. Originally a bond did not itself create a debt, but was merely evidence of an already existing debt. The debt was constituted by a loan of money or other similar transaction, and a deed was executed whereby the debtor acknowledged that he owed the amount. But by virtue of the principle that matter of averment cannot prevail against matter of specialty, the debtor was precluded from going behind his written acknowledgment. He was therefore bound whether the money had been lent to him or not; the bond ceased to be merely evidence of a debt and became a mode of creating one 2. Thus Bracton says: 'An obligation may be constituted by writing; as if a man acknowledge in writing that he owes money, then he is bound by this writing whether any money was actually lent to him or not; nor will he be able to plead against the writing that the money was never paid to him, for he has written that he owes it. A similar instance is the release under seal, which is merely an acknowledgment of payment operating by estoppel as a discharge of the obligation whether there has been performance or not. Similar to the English release is the Roman acceptilatio, which in effect was a method of discharging an obligation, whilst in form it was a solemn acknowledgment of having received performance.

It is scarcely necessary to observe that the principle of estoppel by deed has not been extended to simple writings. An acknowledgment in writing not under seal is not an estoppel but a mere admission; and in the case of unsealed writings the bond has degenerated into the I. O. U. and the release into the receipt. So in Roman law the chyrographum became the cautio and the acceptilatio the apocha 1.

1 Y. B. 21 & 22 Ed. I. 318.

2 "

En dette sur contract le plaintiff monstra in son count pur quel cause le defendant devient son dettour. Autrement in dette sur obligation, car l'obligation est contract in luy meme.' Bellewe, 8 Rich. II. 111 (ed. 1869).

3 Bracton, f. 100 b.

4 It is worthy of note that a deed of feoffment never raised such an estoppel as to

IV. The fourth and last application of the rule that matter in writing must prevail over matter of averment, is found in the doctrine of the inadmissibility of parol evidence to qualify the effect of written instruments. When a contract has been reduced to writing by the parties thereto, the contents of the writing are conclusive evidence of the terms of the contract, and no parol evidence is admissible to vary the contract as expressed in the writing. The law,' says Lord Bacon, will not couple and mingle matter of specialty which is of the higher account with matter of averment which is of inferior account in law 1.' It is not usual to treat this rule as an instance of estoppel, yet it is obvious that the principles are identical. In each case the party is precluded by his acknowledgment in writing from disputing what is so acknowledged. In estoppel however what is so acknowledged is matter of fact, while in the case now under consideration it is the terms of an agreement or other act in law. This rule, like the others that have been discussed, has been established from the earliest times. In 20 Edward I an unsuccessful attempt was made to apply it in excluding evidence that a deed absolute on its face was meant to be conditional on marriage: She ought not to be answered; for see here her deed which witnesses that the gift was absolute and unconditional, and we pray judgment if in opposition to her own deed she ought to be answered. Repeated assertions in this connection that a man cannot deny his own deed meet us in the early reports 3. The rule in question has of course been extended to simple writings.

We have now discussed the various effects of the conclusiveness of matter in writing; it remains to examine the effects of its exclusiveness, the effect, that is to say, of the rule that when matter in writing is available matter of averment is inadmissible instead thereof. It is a familiar maxim that the best evidence must be produced. It is obvious that this maxim can be applied in those cases alone in which some kinds of evidence have obtained legal recognition as being better than others. One of these cases, and indeed the most important of them, is that of written and parol evidence. Consequently where any contract or other act in law has been reduced by the parties thereto to a documentary form, no evidence thereof is admissible save the writing (or in certain cases secondary evidence of its contents). It will be seen later on that exclude an averment that the grantee did not obtain seisin. 'A charter is worth nothing without seisin.' Y. B. 33-35 Ed. I. 50. The allowance of such an estoppel would have constituted a revolution in our early law of real property. See Y. B. 21 & 22 Ed. I. 404.

Maxims of the Law, Reg. 25.

3 See Y. B. 20 & 21 Ed. I. 430.

2 Y. B. 20 & 21 Ed. I. 366.

This rule is of course distinct from the rule that a document is the only admissible evidence of its own contents. The latter rule is another and distinct application of the G

VOL. VI.