Mr Chief Justice Marshall suggested that the case might be brought on if the parties would agree that it should stand as if a judgment had been given by the circuit court on the exceptions. The case, he said, could not be heard on a difference in opinion of the judges of the court, on a motion for a new trial.

The counsel for the plaintiffs and defendants having agreed that the case should stand as suggested by the chief justice, and an agreement in writing to that effect having been filed, the court made the following order.

It is now here by the court considered and ordered, that this cause shall now be heard and decided, as on a writ of error brought after verdict and judgment in the circuit court, on the exceptions which were taken in that court; that the cause shall now proceed, as if judgment had been actually entered in the circuit court for the plaintiffs there, and that the certificate in the case shall be taken, regarded and treated as a writ of error, sued out by the defendants below, on the judgment of the circuit court, and that the question shall be, as in other cases, whether the said judgment ought to be reversed or affirmed: but that this court will reserve its opinion and judgment in this cause till the defendants in the court below shall have sued out a writ of error in this cause to the said circuit court, and filed a return thereto, with a bill of exceptions in this cause, in the usual form, signed by the court below, in this court.

The case came on for argument, after the defendants had sued out a writ of error on a judgment entered in the circuit court for the plaintiffs, in conformity with the suggestion and order of this court. The case was as follows:

The action was brought to recover damages for an alleged infringement of a patent right, and came on for trial in the circuit court in November 1828, when a verdict was rendered for the plaintiffs for three thousand two hundred and sixty-six dollars sixty-six cents.

The plaintiffs gave in evidence a patent from the United States, in the following terms, the same being the patent declared upon:

'Whereas Joseph Grant, a citizen of the United States, hath alleged that he hath invented a new and useful improvement in the mode of manufacturing hat bodies by the combination of motions, viz. the rotatory and revolving motion, with the vibrating or transverse motion, which forms the two hat bodies by machinery, and crosses the wool from one extremity of the hat bodies to the other at one operation, called Grant's improved winding machine for setting up hat bodies—his former patent for the same invention, dated the 11th day of August 1821, having been cancelled, owing to the defective specification on which the same was granted—which improvement he states had not been known or used before his application—hath made oath that he does verily believe that he is the true inventor or discoverer of the said improvement; hath paid into the treasury of the United States the sum of thirty dollars, delivered a receipt for the same, and presented a petition to the secretary of state, signifying a desire of obtaining an exclusive property in the said improvement, and praying that a patent may be granted for that purpose.

'These are, therefore, to grant, according to law, to the said Joseph Grant, his heirs, administrators or assigns, for the term of fourteen years, from the 11th day of August 1821, the full and exclusive right and liberty of making, constructing, using, and vending to others to be used, the said improvement, a description whereof is given in the words of the said Joseph Grant himself in the schedule hereunto annexed, and is made a part of these presents.

'In testimony whereof, I have caused these letters to be made patent, and the seal of the United States to be hereunto affixed. Given under my hand, at the city of Washington, this 28th day of April, in the year of our Lord 1825, and of the independence of the United States of America the forty-ninth.

[L. S.] J. Q. ADAMS.

By the President. H. CLAY, Secretary of State.

City of Washington, to wit:—I do hereby certify, that the foregoing letters patent were delivered to me on the 23th day of April in the year of our Lord 1825, to be examined; that I have examined the same, and find them conformable to law, and I do hereby return the same to the secretary of state within fifteen days from the date aforesaid, to wit, on this 28th day of April, in the year aforesaid.

WILLIAM WIRT, Attorney General of the U. S.

The schedule referred to in these letters patent, and making part of the same, contained a description, in the words of the said Joseph Grant himself, of his improvement in the mode of manufacturing hat bodies by the combination of motions, viz. the rotary or revolving motion, with the vibrating or transverse motion, which forms the two hat bodies by machinery, and crosses the wool from one extremity of the hat bodies to the other at one operation, called Grant's improved winding machine for setting up hat bodies; his former patent for the same invention, dated on the 11th day of August, A.D. 1821, having been cancelled, owing to the defective specification on which the same was granted.

The schedule, which contained a full description of the invention, and of the mode of using it, was also given in evidence.

The counsel for the plaintiffs also produced and read in evidence a certificate of the secretary of state, duly authenticated under his hand and official seal, and certain papers thereto annexed, in the words and figures following:

To all to whom these presents shall come, greeting:

I certify that the annexed is a true copy of the record of cancellation of a patent granted to Joseph Grant on the 11th of August 1821, and cancelled on the 28th of April, A. D. 1825; also, that the annexed is a true copy of the petition praying for the cancellation, and the issuing of another patent for the same invention.

In testimony whereof, I, Henry Clay, secretary of state of the United States, have hereunto subscribed my name, and caused the seal of the department of state to be affixed. Done at the city of Washington, this 19th day of May A. D. 1828, and of the independence of the United States of America, the fifty-second.

[L. S.] H. CLAY.

This patent was returned to the patent office, the seal broken, and now stands cancelled, owing to the defective specification on which it was issued, and another patent granted (with a corrected specification, and for the 28th day of April 1825, bearing date with the first, and for the same invention.

The petition of Joseph Grant, of Providence, in the county of Providence, and state of Rhode Island, hatter, a citizen of the United States of America, respectfully represents, that your petitioner has invented a new and useful improvement in the mode of manufacturing hat bodies by the combination of motions, viz. the rotatory or revolving motion with the vibrating or traverse motion, which forms the two hat bodies by machinery, and crosses the wool from one extremity of the hat bodies to the other at one operation, called 'Grant's improved winding machine for setting up hat bodies,' according to the specification, explanations, and drawings, herewith presented; which the subscriber prays may be taken as a part of his petition—an improvement not used or known before his application, the advantages of which your petitioner is desirous of securing to himself and his legal representatives.

Your petitioner would further state, that he has, heretofore, viz. on the 11th day of August, A.D. 1821, obtained letters patent from the president of the United States for his said improvement, but, owing to a defective specification on which the same were granted, he prays that the said patent may be cancelled, and a new and correct one granted, embracing the same improvements, so far as the same are set forth in the accompanying specification, drawing, and explanations. Your petitioner therefore prays that letters patent of the United States may be issued, granting to your petitioner, his heirs, administrators, or assigns, the full and exclusive right of making, constructing, using, and vending to others to be used, his said improvement, according to the specification and drawings hereto annexed, agreeably to the act of congress in such case made and provided; your petitioner having paid thirty dollars into the treasury of the United States, and complied with other provisions of the said act. As in duty bound, will ever pray.

JOSEPH GRANT.

To the Hon. Henry Clay, Secretary of State of the United States of America.

Providence, April 20, 1825.

As the assignee of David Curtis, who was the assignee of Joseph Grant, of one moiety of the original patent, I unite in the prayer of the above petitioner, that the original patent may be cancelled, and a new one granted to the said Joseph Grant, as above set forth.

SOLOMON TOWNSEND.

To the Hon. Henry Clay, Secretary of State of the United States of America.

Providence, April 20, 1825.

The counsel for the defendant objected that the secretary of state had no power, by law, to accept a surrender of and to cancel said letters patent, or to inquire into or decide upon the causes for so doing, or to grant said second patent for the same invention, with an amended specification, for the unexpended portion of the term of fourteen years, which had been granted by the first patent.

The court decided that such surrender might be made, when the defect in the specification arose from inadvertence or mistake, and without any fraud or misconduct on the part of the patentee; and that the secretary of state had authority to accept such surrender, and cancel the record of the patent, and to issue a new patent for the unexpired part of the fourteen years granted under the old patent in manner aforesaid. To which decision the counsel for the defendant excepted.

The fourth plea, filed on the part of the defendant, after reciting the specification annexed to the patent of the plaintiffs, averred as follows: 'and the defendants aver that said specification does not correctly or accurately describe the improvement claimed by the said Joseph Grant as his invention; but said specification, and the drawings thereto annexed, are altogether defective in this, among other things, namely: in said specification no proportions, sizes, or distances, are given, and the bigness or size of none of the principal parts of said machine is given in said specifications or drawings, but the same is wholly omitted; and, in other particulars, said specification and drawings are altogether defective. And the defendants aver that said specification, annexed to and making part of said letters patent, with the drawings thereto annexed, do not contain a written description of his, the said Joseph Grant's invention and improvement aforesaid, and manner of using it, in such full, clear, and exact terms as to distinguish the same from all other things before known, and so as to enable any person skilled in the art of which said machine or improvement is a branch, or with which it is most nearly connected, to make and use the same; and that, for the cause aforesaid, said letters patent are void. All which the defendants are ready to verify; wherefore they pray judgment if the said Joseph Grant and Solomon Townsend ought to have or maintain their aforesaid action against them, and for their costs.' To which plea the following replication was filed:

And as to the said plea of the said Eliakim Raymond and Henry Raymond, by them fourthly above pleaded, the said Joseph Grant and Solomon Townsend say, that, by reason of any thing in the said last mentioned plea alleged, they ought not to be barred from having and maintaining their aforesaid action thereof against them, the said Eliakim Raymond and Henry Raymond, because they say that the specification mentioned in the said last mentioned plea does correctly and accurately describe the improvement claimed by the said Joseph Grant, as his invention; and because they say further, that neither the said specification, nor the drawings thereto annexed, are defective in any of the particulars in that behalf alleged in the said last mentioned plea; and this they, the said Joseph Grant and Solomon Townsend, pray may be inquired of by the country; and the said Eliakim Raymond and Henry Raymond do the like, &c.

The counsel for the defendants on the said trial introduced sundry witnesses to prove the allegations traversed in the said fourth plea, and insisted that they had proved the same, and that the said specification of the said Joseph Grant did not describe the improvement which he claimed to have invented, in such full, clear, and exact terms as to distinguish the same from all other things before known, nor so as to enable a person skilled in the art or science of which it is a branch, or with which it is most nearly connected, to make and use the same; and they requested the court to charge the jury, that if they found that the defendants had maintained and proved their averments in that respect, that they must find the same for the defendants; which instructions the court refused to give, but instructed the jury that the patent would not be void on this ground, unless such defective or imperfect specification or description arose from design, and for the purpose of deceiving the public; to which opinion the counsel for the defendants also excepted.

The case was argued by Mr Webster, for the defendants in the circuit court, now before the court as plaintiffs in error; and by Mr Ogden, for the defendants, plaintiffs in the circuit court.

Mr Webster stated, that the first question for the decision of the court, was whether the secretary of state can accept the surrender of a patent, cancel it, and grant a new one for the unexpired term for which a patent had been granted, on a suggestion that the specification is defective through inadvertence or mistake?

It cannot but be doubted whether such cancellation can be made any where, so as to take out a new patent.

The whole system of patents rests on statute provision. There is no common law power, or prerogative right, in the president to issue a patent. In this particular, our law is different from the English. Ours is a statute grant; theirs is an emanation out of a statute prohibition. With us, the fountain is statute; with them, prerogative. Our statute makes no provision for any surrender, and the issuing of a new patent thereon. Indeed, it seems impossible to reconcile such a proceeding to the requisitions of the act.

How can the patentee allege, or assign for his second patent, that his improvement had not been known before that application. That is a statute requisition; here, in the case before the court, it had been in use three years. The party must claim, and in this case does claim, under his second patent, as a new and substantive patent; not under his first, with an amended specification.

On surrender or cancellation of the patent, the party comes for a new patent, for the same invention. Now, in all such cases, it must be that the thing has been used or known before that application. In this very case, it does not appear that his petition contained the statute requisites; it refers it to his former application.

But however this may be, the secretary of state has no authority to make a record of cancellation, and to issue a new patent. The secretary of state is a merely ministerial officer. All the laws relating to granting patents regard him as merely ministerial. His department is denominated an executive department. He has no where any particle of judicial power. On the subject of patents, as well as all others, he acts wholly ministerially.

By the first patent law, April 10, 1790, the secretaries of state, war, and attorney-general, were invested with authority to grant or refuse patents. But this was repealed by the existing law of February 21, 1793. The power of granting or withholding patents, was by this act taken away from every body. The secretary was to give the patent out, on certain requisitions being complied with, without exercising any judgment, or making any inquiry; and the patent was to avail the grantee what it might, according to the truth of the representations of the patentee. It is matter of right, and matter of course, to issue the patent; if the requisites of the law be complied with. The secretary has nothing to do but make out the patent.

He is applied to as keeper of the seal. And if a dispute arises between inventors, the secretary cannot decide it; he is to appoint umpires. Sect. 9. He is not trusted to decide even whether the form of the letters patent be conformable to law; that belongs to the attorney-general.

The general power of the secretary is commented on in Marbury v. Madison, 1 Cranch, 159, 160. He is to record diplomatic transactions; but he cannot alter or cancel those records. He records treaties and acts of congress; but those, of course, he cannot alter.

The secretary has no power to record in his office any transaction, not his own, except so far as authorised by statute.

In regard to patents, the statute declares what shall be recorded.

1. The patent itself. 2. The assignment of it, if requested. These are all. The secretary has nothing more to do. As to cancelling a patent, he is functus officio.

The breaking the seal, or other cancellation by the patentee of his patent, is just as effectual if done any where else as if done at the state department. It is not a transaction which the law has prescribed for recording any where. He can neither authorise this cancellation, nor forbid it; nor make it matter of official record.

The vacating and cancelling the record of a patent is in its nature a judicial act. The act of 1793 treats it as such, and provides two modes of such vacation. Cited the 6th and 10th sections of the act.

The common law, also, provides a mode. The court cannot add a new section to the act.

In England, the repealing of patents is always regarded as a judicial act. Godson on Patents, 200.

The limitations, under which the court below thought the secretary could vacate one patent and issue another, show that the power he is expected to exercise is judicial. The defect must arise from inadvertence or mistake, without fraud or misconduct. The power of the secretary to act, then, depends upon his having adjudged the case to be one of mistake, and not a case of fraud. This is judicial power. How can the secretary make this judicial investigation.

1. He has no parties before him, and no power to bring persons before him. When done in court, this is done in the presence of litigant parties. 2. He cannot summon witnesses before him, and if they should come voluntarily, he cannot administer an oath to them. Such oaths would be extrajudicial and nugatory. He cannot require the party's own oath.

Now, whether the want of a proper specification be the effect of accident or of fraud, is a question of fact; and it is to be decided by the secretary, without parties, oaths or witnesses.

Besides, the surrender is to be accepted only when the specification is defective. Is not that very question a question of fact, or a mixed question of law and fact? Certainly it is. It is for the jury: even a court cannot determine it without a jury.

By act of April 20th, 1818, the secretary of state appoints a superintendent. Does he delegate to him his judicial power?There is but one way of answering this view of the subject. It must be contended, that in every case, on the mere suggestion of the patentee, a new patent is to be issued for the residue of the term, with an amended specification: leaving it to be decided, when suit should be brought, whether the defect was inadvertent or fraudulent.

This would change the whole patent system. Its effects would be monstrous. Patentees would try their claims under one specification; they might fail; and they would call it inadvertence, and try another experiment.

A man builds an expensive factory, puts in costly machinery, not patented, not described in any specification; he expends much money; by and by, he is sued for violating a patent, and he finds that since he built, an old patent has come out with a new specification. A hearing, of which he knew nothing, has been had before the secretary, and a new patent has issued, and he is called on to stop his factory. Now this supposed case is the very case before the court, The defendants erected their works in 1823, 1824. They knew of the plaintiff's patent of August 11, 1821, but it did not describe any machinery used by them. But in 1825 he surrendered his first patent, took out another, with a specification describing their machinery, and sued them. Under the direction of the court he has recovered a verdict for three thousand two hundred and sixty-six dollars; and is entitled, of course, to have this trebled, and the defendants are ruined. Is this legal? A bill in equity is pending also, to stop the defendants' factory.

Now, what reason is there for saying that defendants shall suffer these losses, even by the mistake or inadvertence of the patentee. The invention had become public, and if not protected by the first patent, it was gone forever. A bad patent is no patent. 1 Barn. & Ald. 386. It may be well for congress to give the courts or judges power to vacate patents on patentee's own motion; but then congress would provide limitations and securities for innocent persons. On the doctrine of this case there are no securities.

What are the consequences of such a principle? A man finds out there are other machines made like his, which he would like to stop; he sets up inadvertence, gets a new patent, and stops them. If he swears it, who can deny it? It is of younger date, and he swears he meant to describe it. This would furnish irresistible temptation to perjury.

There is a deeper objection. A man makes an invention; he gets a patent, but his specification does not describe it, but describes something else. In the meantime, the public use, not what he has patented, but what he has not. Now, how is the public to be deprived of the use of this? It is denied that they can be prevented this use. The decision of this court in Pennock v. Dialogue, proves this cannot be. The invention was used and known before it was patented. A patent, not describing an invention, is void as to that invention, and does not protect it. The invention, by a single month's use, unprotected by a patent, becomes public property and can never be resumed.

Whether the patent be void through fraud or inadvertence, if it does not describe the invention, then the invention is not protected, but has become public property. Pennock v. Dialogue so decides.

In England a scire facias to repeal a patent is a criminal proceeding, and does not allow costs. 7 Term Rep. 367; Godson, 201.

This case was tried in New York before Pennock v. Dialogue was decided in this court.

The only case applicable to the one now before the court is Morris v. Huntington, 1 Paine, 348. That was decided in 1824. Its doctrine was materially changed by Pennock and Dialogue. After all, that case only decides, that a patent, while another is in existence, is void. The judge then goes on to say, it may be surrendered.

It seems admitted that the new patent ought not to reach back so as to affect those who had already used the invention. But how can this distinction be made? Cited to this point, Brooke v. Clarke, 1 B. and A. 396, note.

The great and conclusive objection is this; the new patent is granted on the new application, and the invention had been public four years. This is fully settled in the case of Pennock v. Dialogue.

The second point is presented on the defendants' fourth plea. The court instructed the jury that the patent would not be void, unless the defective or imperfect specification arose from design, or the purpose of deceiving the public.

This point presents two questions.

1. Whether the patent would be void or not, for the reasons stated; the direction of the court was not pertinent to the issue. The parties were at issue on a question of fact; witnesses were examined, and counsel had summed up; but the court told the jury the issue was immaterial, and under this question the jury found a verdict for plaintiff.

It can require no argument to prove that this is an illegal direction. 2 Day's Reports, 519; 1 Starkie, 388; 9 Cranch, 339, 355.

2. If the question had legally arisen, the law was wrongly stated.

It is insisted, that the plea was a good bar. If the specification was defective, as set forth in the plea, the plaintiff could not recover, whether that defect arose from accident or design. The very words of the third section of the statute require this. There are certain conditions precedent to be complied with, before an inventor can obtain a patent. This is one of them. The language of this section is emphatic and absolute; it could not be stronger. Courts may just as well dispense with the oath.

Suppose he omit to deliver any written specification whatever; can he afterwards say that omission was owing to inadvertence? Yet a defective, or bad specification, is no specification. The same law that requires a written specification, requires a full and accurate one. A defective one is no better than none at all. Suppose he omit to sign the specification, can that be cured?

A party is to describe his invention so as to answer two purposes:

1. To distinguish it from all other things before known. 2. To enable any person skilled in the art to make and use it. If he fail in either of these, he fails in a condition precedent.

This is all very clear, and there would be no doubt about it but for the sixth section. That section has been supposed to raise the doubt.

This section, as has been observed by most judges, is inartfully drawn. It speaks of the right of a defendant to give this act in evidence, in an action founded on the act itself. It is not grammatical. It seems borrowed from the law of 1791.

The first important remark is, that the preceding sections have described the whole extent of the patentee's right. This section has for its object, the giving of immunities and protection to those who may be sued by patentees. Therefore, its object was not to enlarge the right of patentees. Second, most of its provisions are only affirmative, and the right existed before. The action being case, all the material defences are competent. Third, it gives some matters as a defence not mentioned in the first act, such as license or abandonment to the public. Fourth, it does not repeat the same objections to specification.

It allows the defendant to prove three things as fatal defects in his specification. 1. That it does not contain the whole truth relative to the discovery, with intent to deceive the public. This provision may stand with the first section. 2. That it contains more than is necessary to produce the described effect; with intent to deceive the public. This may stand with the third section. 3. That the thing had been used, or described in a public work. This may also stand.

Now, the fraudulent intent is applied only to the two first. In these cases congress may say that fraud shall be proved; because, even with these defects, it may be a patent, and a very useful patent.

It may be a valid patent, though the patentee take it but for part of his invention; yet if he fraudulently deceive the public by keeping part back, it shall be void. So it may be a good and valuable patent, though it contain more than is necessary to produce the described effect. But if it be not so described as to be distinguished from other things before known, or so that skilful persons can use it, it is no patent, or of no use at all. These last objections go to its very existence; and are therefore made prerequisites. They are absolute conditions precedent.

But the main consideration yet is, that this sixth section has an object of its own. It looks, not so much to the defence in its suits, as to the judicial vacation of the patent. It does not look mainly to the defence of the suits, because it leaves out several known grounds of defence. 1. License. 2. Abandonment to the public. 3. That the patent is broader than the invention. 4. That the machine is not well described. All these are defences, and yet not mentioned here.

The object of the sixth section is like that of the tenth, to repeal for fraud proved. All this may stand, without contradicting the third section, or doing violence to its language.

Indeed, Pennock and Dialogue, 2 Peters, 1, has apparently decided this. The sixth section speaks of inventions known before the patentee's discovery. This does not contradict the third section—'known before application.' The court has settled this, and it decides this case.

The plaintiff relied below on the following cases. Park v. Little and Wood, 3 Washington, 196, in April 1813; Gray et al. v. James, 1 Peters's C. C. Rep. 401, 1817.

But it does not appear in this case whether the defendant was or was not proceeding with a view to vacate the patent. Most probably he was; for at that time such was the practice. This is rendered still more probable by a decision looking the other way in 1820, by the same judge. Kneass v. Schuylkill Bank, 4 Wash. C. C. Rep. 13. This case is clearly for the plaintiffs in error.

The case of Whittemore v. Cutter, 1 Gallison, 419, decided in May 1813, was probably a case to vacate; at any rate the judges doubted, and would have divided if the case had turned on this point. The case of Lowell v. Lewis, 1 Mason, 183, in 1817, was probably a similar case; page 189 cited particularly.

How can the court dispense with the express words of the third section. Here they are put into the form of a special plea; can they be disregarded? Suppose there be but one witness; though this is merely a formal requisite, can the court dispense with it?

Besides, why should the statute require a specification, unless it was to be full and accurate? For what purpose should an insufficient specification be enjoined? The court may as well say there shall be none. The statute says as imperatively what the specification shall contain, as that there shall be a specification at all. If an imperfect or defective specification does not render a patent void, what harm does it do to it? Let this question be answered.

Now, the patentee says his specification in 1821 was defective, not through fraud, but defective. But why was not that patent good, or this? If defective through fraud he could not get another. It must be presumed to be defective through inadvertence. If so, what need of a new one?

Compare the two rules of law decided in this case together.

1. The judge held, that party might surrender his patent and take a new one, when the specification in the first was defective through mistake, and without fraud. 2. The court ruled, that though a specification be altogether defective, yet the patent is not void, unless such defect arise from design.

How can these things stand together? If the last proposition be true, all inquiry about the first is idle; unless the question be, whether a man having one good patent may surrender it and take out another good one for the same thing.

It is now matter of settled law, that if a patent be broader than the invention it is void; and it is never inquired whether this arise from design or accident. In nine times out of ten it arises from inadvertence. Now, on what principle is this? Not because the plaintiff's invention has not been invaded. It may have been exactly copied. It is because he has not rightly described what he claimed, and therefore his patent protects him in nothing.

So if a patent be for an improvement of an old machine, it must state the improvement accurately, and distinguish between what is old and what new.

Finally, the English statute has always been construed the other way.

Our sixth section is a substitute for the English scire facias. Cited Davis's Patent Cases, 413; Godson, 124; Holroyd, 100, note.

Cited also Dodson's Patents, 56; 1 Term Rep. 605; 1 Wash. C. C. Rep. 71; 3 Wash. C. C. Rep. 198; 1 Mason, 189, 190; 4 Id. 9, 10; 3 Wheat. 518; 2 Hen. Black. 478; 1 Ves. and Beames, 67; 8 Term Rep. 101; 2 Car. and Paine, 558, 565; 11 East, 107; 14 Ves. 131.

Mr Ogden, for the defendants in error.

The first question is, whether the secretary of state of the United States has a power by law to accept of the surrender of and to cancel a patent which had once been issued, and to grant a second patent for the same invention, with an amended specification for the unexpired portion of the term of fourteen years which had been granted by the first patent?

Upon this question there is not known a single case where the point has been expressly decided in the United States. The patent law is silent upon the subject; and the question must then be decided upon general principles.

A patent for a useful machine is a grant of the exclusive privilege of making and using the machine for a limited time. Now it would seem that a grantee may surrender his grant. A man who has a privilege may surrender that privilege. If a man cancels his patent upon record, it amounts to a surrender of it.

The difficulty in the question, if there be any, must be in the other branch of it. Has the secretary of state the power after the surrender of one patent to grant a new one for the same invention, with an amended specification for the unexpired portion of the term of fourteen years which had been granted by the first patent? Why should he not? When the first patent is cancelled, the invention is unprotected. If a useful one, why should not the inventor have the benefit of it?

He certainly never intended to abandon the benefit of it to the public. His first patent is evidence of that.

A specification requires to be drawn with great accuracy. Mechanics, by whom machines are usually invented, it cannot be supposed are capable of drawing a proper specification. Can it be supposed that the law ever intended to punish their ignorance in drawing a very special legal paper by a forfeiture of all the advantages of their invention?

It is apprehended that the issuing a new patent in England, where there has been no sufficient specification to comply with the condition of the first patent, is pretty much a matter of course.

In the case Ex parte Beck, 1 Bro. Ch. Rep. 575, the lord chancellor says, 'that perhaps upon the petitioner's applying for a new patent, the officers might, under these circumstances, be induced to remit their fees; but that he could give no relief upon the present petition.' Here the lord chancellor speaks of the issuing of a new patent as a matter of course.

In our patent law, the inventor must file his specification before he can procure his patent. In England, the patent contains upon its face a condition that if the patentee shall not make and file a specification within a limited time after the date of the patent, then the patent, and all the liberties and advantages under it, shall cease and be void.

One of the cases cited by the plaintiff in error shows that the specification is not filed within the time mentioned in the patent; although the patent is void, a new one may be taken out by the inventor for the same invention.

This case, in principle, seems to support the proposition for which the defendant in error contends; that if a first patent is void for want of a specification, or for want of a sufficient specification, a new one may be issued for the same invention to the same inventor. In the case of Morris v. Huntington, 1 Paine's Rep. of Cases in the Second Circuit, 355, Mr Justice Thompson says, 'I see no insuperable objection to entering a vacatur of the patent of record in the department of state, if taken out inadvertently, or by mistake. All the proceedings in that department on the subject of patents are ex parte, except in the case of interfering applications. The department acts rather ministerially than judicially, and upon the representation of the applicant, without entering into an examination of the question of right; and there seems to be no good reason why on a like ex parte application, the patent may not be surrendered and cancelled of record, if no misconduct be imputable to the patentee in taking it out.' Cited also page 356.

So far then as this question has ever arisen in any of our courts, the right to surrender an old patent and procure a new one has been recognized. Cited also Barrett et al. v. Hall, 1 Mason, 475, as to the second point.

The second point is settled by the express terms of the patent law, and by the adjudications which have taken place under it.

1. By the express terms of the patent law.

The sixth section of the patent act declares, 'that the defendant shall be permitted to plead the general issue, and give this act and any special matter in evidence, of which notice in writing must have been given, &c. thirty days before the trial; tending to prove that the specification filed by the plaintiff does not contain the whole truth relative to the discovery, or that it contains more than is necessary to produce the described effect, which concealment or addition shall fully appear to have been made for the purpose of deceiving the public, &c. &c.

So, therefore, if the specification does not contain the whole truth relative to the discovery, or if it contains more than is necessary to produce the described effect, the patent shall not, for these causes, or either of them, be evidence unless 'the concealment or addition shall fully appear to have been made for the purpose of deceiving the public.'

2. By the adjudications which have taken place under the law.

The first case in which the question came up, was the case of Whitney v. Carter, in the circuit court of Georgia. That case is cited in Mr Fessenden's Essay on the Law of Patents, 123.

His honour, Judge Johnson, in charging the jury in that case, said, 'he considered the defendants' second objection equally unsupported, and referred to the sixth section of the patent law, by which it is required that the concealment alleged (in order to defeat the patentee's recovery) must appear to have been made for the purpose of deceiving the public. That Mr Whitney could have no motive for such concealment,' &c. &c.

In the case of Gray and Osgood v. James et al. 1 Peters's C. C. Rep. 394, this question came up before the circuit court in Pennsylvania. Judge Washington says, 'but if the jury should be of opinion that the specification is materially defective, the objection will not be sufficient to invalidate the plaintiff's patent, unless they should also be satisfied that the concealment of the circumstances not described, was intended to deceive the public.'

In the case of Whittemore v. Cutter, 1 Gallis. 429, Mr Justice Story says, 'any defect or concealment in a specification to avoid a patent, must arise from an intention to deceive the public.' In the case of Lowell v. Lewis, 1 Mass. 189, the same learned judge makes a similar declaration.

Mr Chief Justice MARSHALL delivered the opinion of the Court.

This action was brought by Grant and Townsend against E. and H. Raymond, to recover damages for an infringement of their right under a patent granted to the plaintiff Joseph Grant, in April 1825. It recited that a former patent had been issued in August 1821, to the same person for the same improvement, 'which had been cancelled, owing to the defective specification on which the same was granted.' The exclusive privilege given by the patent on which the suit is brought, is to continue fourteen years from the day on which the original was issued.

One of the pleas filed by the defendants, contained the following averment: 'and the defendants aver that said specification does not correctly or accurately describe the improvement claimed by the said Joseph Grant as his invention, but said specification, and the drawings thereto annexed, are altogether defective in this, among other things, namely: in said specification no proportion, sizes or distances are given, and the bigness or size of none of the principal parts of said machine is given in said specifications or drawings, but the same is wholly omitted; and in other particulars, said specifications and drawings are wholly defective: and the defendants aver that said specification annexed to and making part of said letters patent, with the drawings thereto annexed, do not contain a written description of his the said Joseph Grant's invention and improvement aforesaid, and manner of using it, in such full, clear and exact terms, as to distinguish the same from all other things before known, and so as to enable any person skilled in the art of which said machine or improvement is a branch, or with which it is most nearly connected, to make and use the same; and that for the cause aforesaid, said letters patent are void.'

The plaintiffs reply that they ought not to be barred, 'because they say that the specification mentioned in the said last mentioned plea, does correctly and accurately describe the improvement claimed by the said Joseph Grant as his invention; and because they say further, that neither the said specification, nor the drawings thereto annexed, are defective in any of the particulars in that behalf alleged in the said last mentioned plea, and this they pray may be inquired of by the country.' On this replication issue was joined.

At the trial the counsel for the defendants objected that the secretary of state had no power by law to accept a surrender of, and to cancel the said letters patent, or to inquire into or to decide upon the causes for so doing, or to grant said second patent for the same invention with an amended specification, for the unexpired portion of the term of fourteen years which had been granted by the first patent.

The court decided that such surrender might be made when the defect arose from inadvertence or mistake, and without any fraud or misconduct on the part of the patentee; and that the secretary of state had authority to accept such surrender, and cancel the record of the patent, and to issue a new patent for the unexpired part of the fourteen years granted under the old patent, in manner aforesaid. To which decision the counsel for the defendants excepted.

After adducing the testimony on which they relied to support their plea herein before stated, the counsel for the defendants moved the court to instruct the jury that if they found that the defendants had maintained and proved their averments in that respect, that they must find the same for the defendants; which instructions the court refused to give, but instructed the jury that the patent would not be void on this ground, unless such defective or imperfect specification or description arose from design, and for the purpose of deceiving the public; to which opinion the counsel for the defendants also excepted.

The jury found a verdict for the plaintiffs, and assessed their damages to three thousand two hundred and sixty-six dollars sixty-six cents: the judgment on which is brought before this court by a writ of error.

The first question in the cause respects the power of the secretary of state to receive a surrender of a patent, cancel the record thereof, and issue a new patent for the unexpired part of the fourteen years for which the original had been granted. The court was of opinion that this might be done 'when the defect in the specification arose from inadvertence or mistake, and without any fraud or misconduct on the part of the patentee.'

The right of the patentee to surrender his patent has not been denied, but the plaintiffs in error insist that no power exists to grant a new patent for the unexpired term. The words of the act, they say, do not confer this power. It cannot be exercised with its necessary guards by the department of state; and inconvenience of no inconsiderable magnitude might result to the public from its exercise. The secretary of state is, in the act of making out patents, a mere ministerial officer, and can exercise no power which is not expressly given.

It is undoubtedly true, that the secretary of state may be considered, in issuing patents, as a ministerial officer. If the prerequisites of the law be complied with, he can exercise no judgment on the question whether the patent shall be issued. It is equally true that the act of congress contains no words which expressly authorise the secretary to issue a corrected patent, if the original, from some mistake or inadvertence in the patentee, should be found incompetent to secure the reward which the law intended to confer on him for his invention. The force of this objection, and of the argument founded on it is felt. If the new patent can be sustained, it must be on the general spirit and object of the law, not on its letter.

To promote the progress of useful arts, is the interest and policy of every enlightened government. It entered into the views of the framers of our constitution, and the power 'to promote the progress of science and useful arts, by securing for limited times to authors and inventors, the exclusive right to their respective writings and discoveries,' is among those expressly given to congress. This subject was among the first which followed the organization of our government. It was taken up by the first congress at its second session, and an act was passed authorising a patent to be issued to the inventor of any useful art, &c. on his petition, 'granting to such petitioner, his heirs, administrators or assigns, for any term not exceeding fourteen years, the sole and exclusive right and liberty of making, using, and vending to others to be used, the said invention or discovery.' The law farther declares that the patent 'shall be good and available to the grantee or grantees by force of this act, to all and every intent and purpose herein contained.' The emendatory act of 1793 contains the same language, and it cannot be doubted that the settled purpose of the United States has ever been, and continues to be, to confer on the authors of useful inventions an exclusive right in their inventions for the time mentioned in their patent. It isthe reward stipulated for the advantages derived by the public for the exertions of the individual, and is intended as a stimulus to those exertions. The laws which are passed to give effect to this purpose ought, we think, to be construed in the spirit in which they have been made; and to execute the contract fairly on the part of the United States, where the full benefit has been actually received: if this can be done without transcending the intention of the statute, or countenancing acts which are fraudulent or may prove mischievous. The public yields nothing which it has not agreed to yield; it receives all which it has contracted to receive. The full benefit of the discovery, after its enjoyment by the discoverer for fourteen years, is preserved; and for his exclusive enjoyment of it during that time the public faith is pledged. That sense of justice and of right which all feel, pleads strongly against depriving the inventor of the compensation thus solemnly promised, because he has committed an inadvertent or innocent mistake.

If the mistake should be committed in the department of state, no one would say that it ought not to be corrected. All would admit that a new patent, correcting the error, and which would secure to the patentee the benefits which the law intended to secure, ought to be issued. And yet the act does not in terms authorise a new patent, even in this case. Its emanation is not founded on the words of the law, but is indispensably necessary to the faithful execution of the solemn promise made by the United States. Why should not the same step be taken for the same purpose, if the mistake has been innocently committed by the inventor himself?

The counsel for the plaintiffs in error have shown very clearly that the question of inadvertence or mistake is a judicial question, which cannot be decided by the secretary of state. Neither can he decide those judicial questions on which the validity of the first patent depends. Yet he issues it without inquiring into them. Why may he not, in like manner, issue the second patent also? The correct performance of all those preliminaries on which the validity of the original depends, are always examinable in the court in which a suit for its violation shall be brought. Why may not those points on which the validity of the amended patent depends, be examined before the same tribunal? In the case under consideration those questions were not supposed by the circuit court to have been decided in the department of state, but were expressly submitted to the jury. The rightfulness of issuing the new patent is declared to depend on the fact, that 'the defect in the specification arose from inadvertence or mistake, and without any fraud or misconduct on the part of the patentee.' The jury were of course to inquire into the face. The condition on which the right to issue the patent depended, could be stated to them for no other purpose.

It has been said that this permission to issue a new patent on a reformed specification, when the first was defective through the mistake of the patentee, would change the whole character of the act of congress.

We are not convinced of this. The great object and intention of the act is to secure to the public the advantages to be derived from the discoveries of individuals, and the means it employs are the compensation made to those individuals for the time and labour devoted to these discoveries, by the exclusive right to make, use and sell, the things discovered for a limited time. That which gives complete effect to this object and intention, by employing the same means for the correction of inadvertent error which are directed in the first instance, cannot, we think, be a departure from the spirit and character of the act.

An objection much relied on is, that after the invention has been brought into general use, those skilled in the art or science with which it is connected, perceiving the variance between the specification and the machine, and availing themselves of it, may have constructed, sold and used the machine without infringing the legal rights of the patentee, or incurring the penalties of the law. The new patent would retro-act on them, and expose them to penalties to which they were not liable when the act was committed.

This objection is more formidable in appearance than in reality. It is not probable that the defect in the specification can be so apparent as to be perceived by any but those who examine it for the purpose of pirating the invention. They are not entitled to much favour. But the answer to the objection is, that this defence is not made in this case; and the opinion of the circuit court does not go so far as to say that such a defence would not be successful. That question is not before the court, and is not involved in the opinion we are considering. The defence when true in fact may be sufficient in law, notwithstanding the validity of the new patent.

It has been also argued that the new patent must issue on the new specification, and on the application which accompanies it. Consequently, it will not be true that the machine was 'not known or used before the application.'

But the new patent, and the proceedings on which it issues, have relation to the original transaction. The time of the privilege still runs from the date of the original patent. The application may be considered as appended to the original application; and, if the new patent is valid, the law must be considered as satisfied if the machine was not known or used before that application.

It has been urged that the public was put into possession of the machine by the open sale and use of it under the defective specification, and cannot be deprived of it by the grant of a new patent. The machine is no longer the subject of a patent.

This would be perfectly true, if the second patent could be considered as independent of the first. But it is in no respect so considered. The communication of the discovery to the public has been made in pursuance of law, with the intent to exercise a privilege which is the consideration paid by the public for the future use of the machine. If, by an innocent mistake, the instrument introduced to secure this privilege fails in its object, the public ought not to avail itself of this mistake, and to appropriate the discovery without paying the stipulated consideration. The attempt would be disreputable in an individual, and a court of equity might interpose to restrain him.

It will not be pretended that this question is free from difficulty. But the executive departments, it is understood, have acted on the construction adopted by the circuit court, and have considered it as settled. We would not willingly disregard this settled practice in a case where we are not satisfied it is contrary to law, and where we are satisfied that it is required by justice and good faith.

We will now proceed to the second exception.

The plea assigns the particular defect supposed to exist in the specification, and then proceeds to aver, in the very words of the act, that it 'does not contain a written description of his the said Joseph Grant's invention and improvement aforesaid, and manner of using it, in such full, clear, and exact terms as to distinguish the same from all other things before known, and so as to enable any person skilled in the art, &c. to make and use the same,' &c.

The plea alleges, in the words of the act, that the pre-requisites to the issuing a patent had not been complied with.

If the matter alleged in this plea constituted no bar to the action, the plaintiffs might have demanded, and have submitted the question of law to the court. But they have chosen to deny the facts alleged in the plea, and to aver in their replication 'that neither the specification nor the drawings thereto annexed, are defective in any of the particulars in that behalf alleged.' Issue was joined upon this replication, and it is that issue which the jury were sworn to try.

At the trial the counsel for the defendants, after the evidence was closed, asked the court, in substance, to instruct the jury that if they should be of opinion that the defendants had maintained and proved the facts alleged in their plea, they must find for the defendants. The court refused this instruction. Ought it to have been refused? If in the opinion of the jury the defendants have proved maintained every fact alleged in the plea on which the issue they are sworn to try is joined, ought not the jury to find that issue for the defendants? Is not this required by their oaths? The conclusion, 'and that for the cause aforesaid, said letters patent are void,' is an inference of law from the facts previously alleged; not the allegation of a distinct fact to be submitted to the jury.

The court proceeded to instruct the jury 'that the patent would not be void on this ground, unless such defective or imperfect specification or description arose from design or for the purpose of deceiving the public.'

Now this 'design,' this 'purpose of deceiving the public,' constituted no part of the issue. The defendants had not alleged it, and could not be supposed to come prepared to prove it. A verdict for them would not imply it. The instruction is understood to direct a verdict which finds in fact that the description or specification is not defective; and this verdict against the evidence is to be found because that defect 'arose not from design, or for the purpose of deceiving the public.'

But we must inquire whether the instruction, independent of its departure from the issue, be consistent with law. It is 'that the patent would not be void unless,' &c.

The fifth section of the act gives the party aggrieved an action for the infringement of his patent right. The sixth provides 'that the defendant in such action shall be permitted to plead the general issue, and give this act in evidence, and to give in evidence any special matter, of which notice in writing may have been given to the plaintiff or his attorney thirty days before trial, tending to prove that the specification filed by the plaintiff does not contain the whole truth relative to his discovery, or that it contains more than is necessary to produce the described effect; which concealment or addition shall fully appear to have been made for the purpose of deceiving the public; or that the thing thus secured,' &c. 'in either of which cases judgment shall be rendered for the defendant with costs, and the patent shall be declared void.'

Courts did not, perhaps, at first distinguish clearly between a defence which would authorise a verdict and judgment in favour of the defendant in the particular action, leaving the plaintiff free to use his patent, and to bring other suits for its infringement; and one which, if successful, would require the court to enter a judgment not only for the defendant in the particular case, but one which declares the patent to be void. This distinction is now well settled.

If the party is content with defending himself, he may either plead specially; or plead the general issue, and give the notice required by the sixth section of any special matter he means to use at the trial. If he shows that the patentee has failed in any of those pre-requisites on which the authority to issue the patent is made to depend, his defence is complete. He is entitled to the verdict of the jury and the judgment of the court. But if, not content with defending himself, he seeks to annual the patent, he must proceed in precise conformity to the sixth section. If he depends on evidence 'tending to prove that the specification filed by the plaintiff does not contain the whole truth relative to his discovery, or that it contains more than is necessary to produce the described effect,' it may avail him so far as respects himself, but will not justify a judgment declaring the patent void, unless such 'concealment or addition shall fully appear to have been made for the purpose of deceiving the public;' which purpose must be found by the jury to justify a judgment of vacatur by the court. The defendant is permitted to proceed according to the sixth section, but is not prohibited from proceeding in the usual manner, so far as respects his defence; except that special matter may not be given in evidence on the general issue unaccompanied by the notice which the sixth section requires. The sixth section is not understood to control the third. The evidence of fraudulent intent is required only in the particular case, and for the particular purpose stated in the sixth section.

This instruction was material if the verdict ought to have been for the defendants, provided the allegations of the plea were sustained, and if such verdict would have supported a judgment in their favour; although defect in the specification might not have arisen from design, and for the purpose of deceiving the public. That such is the law we are entirely satisfied. The third section requires, as preliminary to a patent, a correct specification and description of the thing discovered. This is necessary in order to give the public, after the privilege shall expire, the advantage for which the privilege is allowed, and is the foundation of the power to issue the patent. The necessary consequence of the ministerial character in which the secretary acts, is that the performance of the pre-requisites to a patent must be examinable in any suit brought upon it. If the case of the first impression, we should come to this conclusion; but it is understood to be settled.

The act of parliament concerning monopolies contains an exception on which the grants of patents for inventions have issued in that country. The construction of so much of that exception as connects the specification with the patent, and makes the validity of the latter dependent on the correctness of the former, is applicable, we think, to proceedings under the third section of the American act. The English books are full of cases in which it has been held that a defective specification is a good bar when pleaded to, or a sufficient defence when given in evidence on the general issue, on an action brought from the infringement of a patent right. They are very well summed up in Godson's Law of Patents, title Specification: and also in the chapter respecting the infringement of patents: also in Holroyd on Patents, where he treats of the specification, its form and requisites. It is deemed unnecessary to go through the cases, because there is no contrariety in them, and because the question is supposed to be substantially settled in this country. Pennock and Sellers v. Dialogue, 1 Peters, 1, was not, it is true, a case of defect in the specification or description required by the third section, but one in which the applicant did not bring himself within the provision of the first section, which requires that before a patent shall issue the petitioner shall allege that he has invented a new and useful art, machine, &c. 'not known or used before the application.'

This pre-requisite of the first section, so far as a failure in it may affect the validity of the patent, is not distinguishable from a failure of the pre-requisites of the third section.

On the trial evidence was given to show that the patentee had permitted his invention to be used before he took out his patent. The court declared its opinion to the jury, that if an inventor makes his discovery public, he abandons the inchoate right to the exclusive use of the invention. 'It is possible,' added the court, 'that the inventor may not have intended to give the benefit of his discovery to the public.' 'But it is not a question of intention,' 'but of legal inference, resulting from the conduct of the inventor, and affecting the interests of the public. It is for the jury to say whether the evidence brings this case within the principle which has been stated. If it does, the court is of opinion that the plaintiff is not entitled to a verdict.'

The jury found a verdict for the defendants, an exception was taken to the opinion, and the judgment was affirmed by this court.

This case affirms the principle that a failure on the part of the patentee, in those pre-requisites of the act which authorise a patent, is a bar to a recovery in an action for its infringement; and that the validity of this defence does not depend on the invention of the inventor, but is a legal inference upon his conduct.

Upon these authorities and this reasoning we are of opinion that the instruction was erroneous, and that the judgment ought to be reversed and the cause remanded. One of the judges composing the majority thinks, that the direction would have been erroneous on a plea properly framed upon the third section of the act, and averring the facts of a defective specification, and a non-compliance with other requisitions of that section, for that such a plea would be a good bar and defence to the action: but, in his view, the plea relies upon the facts as avoiding the patent entirely, and avers it to be void. He thinks, however, that the replication puts the facts, and not the point whether void or not, in issue; and that the direction of the court was erroneous, since it was equivalent to a declaration, that if all the facts were proved, the issue ought not to be found for the defendants, unless the imperfection of the specification arose from a fraudulent design.

The judgment is reversed, and the cause remanded with directions to issue a venire facias de novo.

This cause came on to be heard on the transcript of the record from the circuit court of the United States for the southern district of New York, and was argued by counsel. On consideration whereof, it is the opinion of this Court that the said circuit court erred in instructing the jury 'that the patent would not be void on this ground, unless such defective or imperfect specification or description arose from design, and for the purpose of deceiving the public.' Whereupon it is ordered and adjudged by this Court, that the judgment of the said circuit court in this cause be, and the same is hereby reversed, and that this cause be, and the same is hereby remanded the said circuit court, with directions to award a venire facias de novo.