in the name of the principal, it is not sufficient. On the other hand, if the agent appears in the body of the deed to act for himself, but executes it in the name of his principal, it will be valid.

Fifthly. As a power of attorney, in se, vests no interest or estate in the agent, none can pass from him, nor can any pass through him, unless in the name of the principal, who hath the estate. Hence the courts have distinguished between agencies relating to mere acts, and those which authorize the passing of an interest. In both cases, however, the act ought to be in the name of the principal, but, in the former case, it requires less to make it the act of the principal, than in the latter. Where a mere act is to be done, it may be stated to be done for and in behalf of the principal; but where an interest is to be conveyed, the name of the principal must be signed by the agent, and the seal must be appropriated by the agent for him. It was on this very distinction that Combes's celebrated case turned. 9 Coke, 77.

Sixthly. The principles of law must be respected, even though hardship and inconvenience do sometimes flow from their application. Nor can this rule of law be dispensed with, even by the consent of the principal. The concluding words, therefore, of Mullins's power, viz. 'and any instrument of writing given by my attorneys, shall be as binding on me as if I were personally present,' are wholly inoperative to sustain any deed, illegally executed, independently of that clause. But this clause (as I have before stated) must be connected with the previous sentence of the instrument, viz. that which covenants to ratify all that his attorneys may lawfully do.' But even admitting that the last clause is not thus controlled, still, as I have said, it cannot legalize any instrument, however informal; nor can it give efficacy to a deed or validate that, as a due execution of a power, which law pronounces to be an undue execution. I may go still further in this doctrine. Suppose that Mullins had expressly stated that 'whereas by a rule of law an agent must execute a deed in the name of his principal, in order to convey the principal's property to a purchaser, he, John Mullins, authorized his said agents to make the deeds in their own names, hereby ratifying and confirming, as duly executed deeds of conveyance, any deed made by his said agents.' Such a deed, I apprehend, would not per se be valid to convey the land; first, because it would perhaps be in

disregard of the act of 1785, and secondly, it would certainly come in conflict with the principles of law, which no consent can dispense with. Even under this strong view of the case, the deed, under which the defendants claim, would be inoperative as a conveyance, and the matter would rest in covenant only; a multo fortiori the present power cannot have that effect, when it is obvious, from the whole instrument, that no such intention of dispensing with any rule of law is manifested by the instrument. The doctrine that explicit declarations of consent will not give efficacy to instruments beyond what is accorded to them by the rules of law, is perhaps too plain to have required any argument; it is exemplified by a great variety of cases and analogies. Hence is it that the rule in Shelly's case must operate where the premises exist on which it may operate, although there be the most explicit intention that the rule shall not operate. So also, if the legal operation of technical words in an instrument be to create a certain estate, they must so operate, though contrary to the clearest intention of the conveyance or even express words. This is often the case, although in deeds, as well as in writs, the general rule of law professes to regard intention.

Seventhly. Regrets have been expressed by some that the rule of law now under consideration should have been established or be continued. They regard it as too technical and too liable to be infringed, to be sound. It originated, however, in principle and good sense. It has become a rule of property, and, like some other doctrines of the law, may occasionally operate with severity. But who is competent to rescind a clear and established legal rule or doctrine? No other power, we presume, than the legislature. If courts can do so, where is the safety or certainty of the science? Judicial legislation might riot on the labors of past ages, and the very foundations of a science, which ought to be as fixed as the eternal mountains, until altered by the legislature for good causes, would be liable to continual change, and thus, perhaps, be productive of more hardship than any fixed or certain rule. In maintaining this position I am not so fast a friend to the doctrine of stare decisis as not to admit some latitude to judicial interpretation; but when the rule is fixed by long concurring adjudication, it is nothing less than legislation of a very dangerous kind, for courts to annul or qualify the rule, when it happens to prove inconvenient or even unphilosophical. It is the province of

the legislature to abolish it when it proves a general evil; and so far from the legislature of Georgia having expressed any dissatisfaction at the rule, by the act of 1785, it appears to have left the execution of such sales and conveyances, under powers of attorney, as it is at common law. We are not, however, disposed to regard the rule in question as too technical for practical purposes. There are very many other rules more difficult to be thought of or adhered to, and the courts have never considered it as an arbitrary doctrine, but have almost uniformly enforced it with alacrity in every case in

which it has occurred.

Eighthly. But it has sometimes been objected that attorneys often execute in both ways, and that in some places they always sign their own names as attorneys for their constituent, and therefore that the deed ought to be permitted by the courts to operate, on the ground that communis error facit jus. In reply to this idea, I beg leave to state that errors merely acquiesced in, and never judicially litigated and decided, do not constitute the law of a particular case, under the auspices of this maxim. The practice under the Lords's act, for example, had been one way for more than thirty years; and yet in the case of Lench v. Pargiter, Doug. 68, (which was the first case in which that statute came under judicial notice) the practice was declared to be incorrect, and that the maxim communis error facit jus would not apply. So also, in nearly every state of the Union as well as in England, it is a general practice to regard an interest in a pew, or shares in a road or canal company, a rail road company, bridge company, and finally the stock of all companies deriving revenue mainly from the use of corporeal hereditaments as personal, and not as real estate, or incorporeal hereditaments, annexed to, and exercisable in the realty; and yet it has been decided, and correctly we think, that such estate is real and not personal. It has been held that the shares descend to the heir, and not to the executor; that a wife may be endowed of them, a husband have curtesy, and finally that they possess every attribute of real estate. In this instance we find that long practice does not make the law. Hence it has become, of late, quite usual to express in the charters of such companies that the stock shall be deemed personal estate; lest, without this statutory provision, it would descend as other real estate. Vide Buckeridge v. Ingram, 2 Ves. Jr. 652. Many other cases might, but

need not be cited. And likewise, as another example, we may cite Way v. Carey, 1 New York T. R. 191, it was held that, although a justice's court, ever since the revolution, had entertained jurisdiction of suits by and against executors and administrators, yet when the question was made, the Supreme Court had no hesitation in declaring that long practice had no effect, and that the justice's court had no such power. So, lastly, where the corporation of New York had been in the practice of creating penalties, and giving a moiety of them to the informer, when the validity of the practice was questioned, the court wholly rejected the maxim, communis error facit jus, and denied the right, although it had been exercised for forty years, and very many penalties had, during that period, been even judicially recovered. Vide 1 Caines Cases in Error, 54.

Admitting, then, that in Georgia, or elsewhere, the usual practice has been for agents or attorneys to execute the deeds in disregard of the rule we are considering, the deed in question can derive no support from an appeal to the maxim we have remarked on.

Ninthly. The books are full of cases referrible to two classes or heads, both of which involve the question of the necessity of executing the deed in the name of the principal. The first class includes all those which are strictly applicable to the deed in controversy, and the point whether the principal is bound by a deed made by an attorney in his own name, as attorney for the principal; and the second class embraces those cases in which the inquiry is, whether the agent has made himself personally responsible, by reason of his neglect to execute the deed in the name of his principal. We should take care not to confound these two classes of cases.

Tenthly. The rule has no exceptions that I am aware of. For though it does not apply to simple contracts made under a general authority, (though even in these cases the agency must appear) nor to the case of a sale made under a power given by will, the devisor being dead, the person authorized may sell in his own name; yet these are not so much exceptions to the rule, as cases in which the premises, on which the rule is based, do not exist. 9 Coke, 77 a; 1 Roll's Abr. 330, plac. 40.

The foregoing preliminary observations will considerably diminish our investigation of the cases which sustain the doctrine of the invalidity of Young and Blake's deed, (admitting

a right to convey) on account of its irregular execution under

the power.

I now proceed to an examination of the cases. The earliest recorded case, which we need notice, is that reported by Moore, page 70, plac. 191, and which occurred 6 Elizabeth. I shall render the case into English. The king, by his patent, gave authority to his surveyor to make leases of certain lands for life, rendering the ancient rent. The surveyor made the indenture thus, "Hæc indentura facta inter Dominum Regem, ex una parte, et Jacobum Southard, ex altera parte, quod Rex Dominus dimisit," &c. in testimony of which the surveyor attached his own seal. And by the opinion of the judges the lease was held to be void; for how could a man give an interest by deed of grant, but so long as he well used his authority; for the servant cannot make leases from year to year, in his own name, but must do so in the name of his master. Thus the seneschal of a manor, of which there are many copyholders, cannot make the copies in his own name, but must say Dominus per senescum suum concessit. So also he should

not have put his own seal to the lease, but the seal of the king, because it was not a lease from the king without his seal. And the words prove this, as he should say the king by A B hath attached his seal. There was a great case in Devonshire, where a bailiff having made such a lease by virtue of such a power, in his own name. It is the opinion of the court, therefore, that all such leases are void.'

The above is certainly a very rudely reported case, as most of the cases of those early times are. I admit that in this case the attorney or surveyor sigillum suum apposuit merely, and did not add to his seal and signature attorney for the king.' But this can make no difference, (as all subsequent cases abundantly prove) 1. Because the infirmity consisted in placing his own seal and signature, which could not have been cured by the addition of the words, attorney for the king.' 2. The deed itself said Dominus Rex dimisit, and yet this (though it showed the agency) did not make it the king's deed, through the ministry of his surveyor, because of the irregularity in not executing the deed in the name of the king.

I am fully aware that many attempts have been made to distinguish the case of a mere seal and signature by the agent, from those cases in which the agent has added to his name