Dyer, J.
The bond was prepared April 30, 1874, in the office of the collector of internal revenue. The written part of the instrument is in the handwriting of one Sherman, who at that time was a deputy in the office. As originally drawn, the names of John M. Stowell, Patrick Walsh, and Hugh P. Reynolds, with their respective residences, were written in the body of the bond. This makes it manifest that the collector understood that Hugh P. Reynolds was to sign the bond as one of the sureties. The bond was signed, as thus drawn, by O’Neill, Stowell, and Walsh, in the collector’s office, on the day of its date. The testimony satisfactorily shows that it was the distinct understanding between O’Neill, Stowell, and' Walsh that Hugh P. Reynolds should be a co-surety on the bond; and I think it was competent for the defense to show this, in view of the fact that the face of the bond as drawn by the collector indicated that Hugh P. Reynolds was to sign the bond as one of the sureties, and that this must have been so understood by the collector. There is a dispute upon the question whether the bond, after its execution by O’Neill, Stowell, and Walsh, remained in the custody of the collector, in expectation that Hugh P. Reynolds would come in and sign it, or whether O’Neill was permitted to take the bond away for the purpose of getting Reynolds’ signature thereto. It seems most probable that the collector retained the custody of the bond; but whether this be so or not, is not in my opinion very material. At all events, there was such delay in procuring the signature of Hugh P. Reynolds—in consequence, as the testimony tends to show, of his absence—that the collector became urgent in his requirement that the execution of the bond by a third surety should be completed. Thereupon O’Neill proposed to the col[569] lector that John B. Reynolds should be substituted as a surety in place of Hugh P.; and upon the representation of O’Neill that John 13. Reynolds was as responsible, pecuniarily, as Hugh P., and that the other sureties would be satisfied with the proposed substitution, the collector caused the word and letter “Hugh P.,” where they occurred in the body of the bond before the name Reynolds, and the residence of that person as written in the bond, to be erased, and substituted therefor the name of John 13. Reynolds, and a description of his residence. Thereupon John B. Reynolds signed the bond as the third surety, and the testimony tends to show that this was done on the twenty-fifth day of Juno, 1874. Of this erasure in the bond, and substitution of John B. Reynolds for Hugh P. Reynolds, the proofs positively show the defendants Stowell and Walsh knew nothing until this suit was begun in 1876. Thus it appears that when Stowell and Walsh signed the bond they understood and expected that Hugh P. Reynolds was to be a co-surety with them; that it must have been also so understood by the collector, because he had drawn the bond accordingly; that subsequently, without consulting Stowell and Walsh, and without their knowledge, the collector, by arrangement with O’Neill, made the change in the bond and permitted the substitution of sureties, which have been stated. Was not this such an alteration of the bond, and such an unauthorized deviation from the original understanding of all the parties, as precludes a recovery against Stowell and Walsh? I am of the opinion that it was.
On the back of the bond there purports to he an acknowledgment of the execution of the bond by all the parties,—O’Neill, Stowell, Walsh, and John B. Reynolds,—dated June 25,1874, before Sherman, deputy collector. If this acknowledgment was in fact taken, it must have been after John B. Reynolds signed the bond, and in that case Stowell and Walsh would be clearly precluded from objecting to the substitution of John B. Reynolds for Hugh P., and to the change in the body of the bond, because it would then be a conclusive presumption that they knew or ought to have known at the time of the acknowledgment of such substitution and change. But both Stowell and Walsh testify with groat positiveness that they never acknowledged the execution of the bond. Their testimony upon that point is not overcome by any proof to the contrary on the part of the government. Sherman cannot be sworn because of mental incapacity. The testimony of the collector, so far as it was thought competent for him to speak upon the subject, is not adequate to meet the positive affirmations of Stowell and Walsh.
The certificate of acknowledgment is not'conclusive, hut only prima facie evidence of what it states. It may be shown to be untrue. Of course, the evidence to overcome it should be strong and convincing. “While a certificate of acknowledgment to a conveyance establishes a prima facie case that the signature of the person purporting to have executed the conveyance is genuine, this presumption will not prevail [570] against positive evidence to the contrary.” Borland v. Walrath, 33 Iowa, 130. See, also, Paxton v. Marshall, 18 Fed. Rep. 361.
The general proposition of law in relation to the liability of sureties laid down by Mr. Justice Story, in Miller v. Stewart, 9 Wheat. 703, is elementary. He says:
“Nothing can be clearer, both upon principle and authority, than the doctrine that the liability oí a surety is not to be extended by implication beyond the terms of his contract. To the extent, and in the manner, and under the circumstances pointed out in his obligation, he is bound, and no further. It is not sufficient that he may sustain no injury by a change in the contract, or that it may even be for his benefit. He has a right to stand upon the very terms of his contract, and if he docs not assent to any variation of it, and a variation is made, it is fatal.”
There is a class of cases, many of which have been cited by the learned counsel for the government, in which it is held that a bond,, perfect on its face, apparently duly executed by all whose names appear thereto, purporting to be signed and delivered, and actually delivered without a stipulation, cannot be avoided by'the sureties upon the ground that they signed it on a condition that it should not be= delivered unless it was executed by other persons who did not execute it, where it appears that the obligee had no notice of such condition,, and there was nothing to put him upon inquiry as to the manner of. its execution, and that he had been induced upon the faith of such bond to act to his own prejudice, Dair v. U. S. 16 Wall. 1; Tidball v. Halley, 48 Cal. 610; State v. Peck, 53 Me. 284; Cutler v. Roberts, 7 Neb. 4; Nash v. Fugate, 24 Grat. 202; Millett v. Parker, 2 Metc. (Ky.) 608; State ex rel. v. Pepper, 31 Ind. 76. Then there are other-cases in which it has been decided that if a bond be written as if to be executed by two or three or more sureties, and it is in fact executed by only one, and is then delivered to the pbligee, it is valid and effectual against that one. Cutter v. Whittemore, 10 Mass. 442. In Russell v. Freer, 56 N. Y. 67, M., plaintiff’s intestate, held the office of collector of internal revenue. Proposing to appoint 0. as his deputy, he required security that C. would pay over all moneys collected, etc.. For this purpose a bond was prepared, which was executed by H. and F., and delivered' to C. When they signed it the name of J. appeared as obligor in the bond, and they were told by O. before signing that J.. would sign it also, and they signed with this expectation. The name of J. was subsequently stricken out of the bond without their knowledge or consent, and it was delivered to M., who had no knowledge of the facts, and who thereupon appointed C. deputy. In an action on the bond, held that H. and F., having placed it in the power of C. to-deliver the bond as a valid and complete instrument, it having been so delivered, and M., having incurred responsibility relying thereon, it was valid and binding.
As will be seen, none of the eases cited meet the facts of the case at bar. Here the conclusion must be, from the manner in which the-transaction took place, that it vas the understanding of all parties,. [571] the collector included, when Stowell and Walsh signed the bond, that Hugh P. Reynolds should sign it as a co-surety. As before observed, the bond was so prepared in the collector’s office, and such was the expectation when Stowell and Walsh signed it, and left it with the collector. The collector had notice of the understanding of the parties. It was not the case of a delivery of the bond with a private agreement between the obligor and the sureties that others should sign it,—an agreement unknown to the obligee. It was not the case of a bond in the hands of an obligor with other names written therein, and then delivered by him absolutely to the obligee, signed by some and not by others. It is not like the case in 56 N. Y. Here the bond was confessedly yet incomplete after Stowell and Walsh signed it, and while it was in the hands of the collector; and through the active instrumentality of that officer or his deputy, and by agreement between him and the obligor, without the knowledge or consent of Stowell and Walsh, the erasure was made in the bond, and a new surety substituted for the one whose name ivas originally written therein, and whom all parties originally expected and understood would sign it as a co-surety. Upon this state of facts I feel obliged to conclude that the bond is not an obligation binding upon Stowell and Walsh.
In Smith v. U. S. 2. Wall. 219, Mr. Justice Gliffobd states the rule to be that any variation in the agreement to which the surety has subscribed, which is made without the surety’s knowledge or consent, and w'liich may jrejadice him, or which may amount to a substitution of a new agreement for the one he has subscribed, will discharge the surety, upon the principle of the maxim non hcec in feeder a veni. And of this case it may be observed that in its facts and upon the law it is highly instructive as bearing upon the kindred question involved in the case at bar.
Several points are made impugning the validity of the assessment described in the complaint, and offered in evidence. The assessment list was for the month of November, 1875, and bears date December 18th of that year. It is contended that in making the assessment the commissioner exceeded his authority in this: that by .section 3182 of the Revised Statutes he was limited in making an assessment against the defendant O’Neill to a period 15 months anterior to the date of assessment; that therefore he could not go back of September 18, 1874; whereas, he did in fact extend the assessment back to May 1, 1874. I do not understand section 8182 as thus limiting the time for making the assessment here in question. By that section the commissioner is first given general power to make the inquiries, determinations, and assessments of all taxes and penalties imposed by title 35 of the statutes relating to internal revenues, and he is required to “certify a list of such assessments when made to the proper collectors, respectively, who shall proceed to collect and account for the taxes and penalties so cer[572] tified.” Then the section provides that whenever it is ascertained that any list which has been or shall be delivered to any collector— that is, any list of assessments already made and certified by the commissioner to a collector, and such as is just before-spoken of—is imperfect or incomplete in consequence of the omission, etc., the commissioner may, at any time within 15 months from the time of the delivery of the list to the collector as aforesaid,—that is, within 15 months after the delivery of the list by the commissioner to the collector,—enter on any monthly or special list the name of such person omitted, etc., and he shall certify and return such list to the collector as required by law. It is observable that this statute does not forbid a reassessment for a period 15 months back of the time when such reassessment is made, but when an assessment lias been made on discovery of an omission, etc., the commissioner may, within 15 months after such assessment, enter on any monthly or special list the name of the person previously omitted. This is what I understand the statute to mean, and the court cannot say, upon the facts before it, that the special taxes against O’Neill here in question, and appearing on the monthly list of November, 1875, or any part of them, were assessed at a time more than 15 months subsequent to any previous list or assessment that may have been imperfect or incomplete from any cause mentioned in the statute.. But it is immaterial, for the purposes of this case, whether I am correct in my interpretation of this provision of the statute or not; for the assessments in question were undoubtedly made under the provisions of section 3253, Bev. St., which declares that “the tax upon any distilled spirits removed from the place where they were distilled, and not deposited in bonded warehouse as required by law,'shall, at any time when knowledge of such fact is obtained by the commissioner of internal revenue, be assessed by him upon the distiller of the same,” etc.
The validity of the assessment is further questioned on the ground that an erroneous rate was adopted by the commissioner in imposing the tax of $20,608.28.on 29,440.40 gallons of distilled spirits from December 1, 1874, to and including March 3, 1875. The tax imposed was at the rate of 70 cents per gallon. On the third day of March, 1875, an act was approved and became the law, changing the rate of tax on distilled spirits to 90 cents per gallon. 18 St. at Large, 618, pt. 3, c. 127. The argument is that this act took effect at midnight of March 2d, and therefore that a tax imposed on spirits distilled March 3d at the rate of 70 cents per gallon was illegal, and that this illegality as to spirits made on that day vitiates the entire assessment. The point thus made is not without force. The question respecting the punctum temporis when a statute takes effect is often one of difficulty; but it would seem 'that the act of March 3, 1875, changing the rate of the tax from 70 cents to 90 cents per gallon, took effect and was in force from the first moment of that day. Arnold v. U. S. 9 Cranch, 104; In re Welman, 20 Vt. 653; In re [573] Howes, 21 Vt. 619. So that, as to spirits produced on the third day. of March, the assessment should have been at the rate of 90 cents, instead of 70 cents, per gallon. Nevertheless, I am not prepared to hold that this vitiated the entire assessment which extended back to December 1, 1874. The defendant O’Neill was not prejudiced by the fact that for one day he was not assessed at as high a rate as the law in force on that day authorized. I do not, therefore, see how he can complain of the alleged irregularity. If liable at all, he was liable to pay 90 cents per gallon on account of spirits produced March 3d, and he was required by the assessment to pay only 70 cents for that day’s production. At most, there was an omission on the part of the commissioner to comply with the full requirement of the law, so far as his act embraced the single day in question, but his action in that respect was not wholly ultra vires. I cannot, therefore, hold that the assessment was invalidated by the act of the commissioner complained of.
The validity of the assessment is further attacked on the ground that a tax of $10,740.80 was imposed on spirits distilled between May 1 and December 31, 1874, and that another tax of $20,608.28 was imposed on spirits produced between December 1, 1874, and March 3, 1875, thus, as it is claimed, making a double tax on the same spirits for the month of December, 1874. But this objection is untenable, because the court cannot say that the two assessments, for the month of December covered the same spirits. Presumably they did not, and if it is a case of double assessment, it is for the defendant affirmatively to show it. The court can by no means presume, in the absence of proof, that the two assessments for the month of December covered the same spirits. It was said on the argument that it was impossible to separate from the property assessed the second time that which had been already assessed once, and which was therefore exempt from taxation. But this assumes, in the absence of proof, that the same spirits were assessed twice, and this assumption is not, in the opinion of the court maintainable.
Concerning that part of the assessment which embraces spirits alleged to have been produced between March 4, 1875, and June 30th of that year, and amounting to $27,786.02, the court does not see how it can be included here as part of the basis of liability upon the bond in suit. The bond expired May 1, 1875. Of course, it only covered transactions occurring between May 1, 1874, and May 1, 1875. The assessment just spoken of, as will be seen, covers a period extending beyond the life of the bond, namely, May and June, 1875. That assessment, covering the period from March 4 to June 30, 1875, is not under the proofs before the court, separable. That is, it is impossible, upon any facts shown here, to correctly and justly determine what, if any, proportion of the spirits produced during that period was so produced and removed during the life of the bond. Perhaps some proportion could be mathematically ascertained on the [574] «basis of the whole amount alleged to have been produced and the number of months and days embraced in the period covered by the ■assessment. But that would be a calculation in its nature arbitrary, ■and might be wholly incorrect, and therefore very unjust. Liability on the bond in suit cannot, therefore, be based upon that assessment.
In the assessment list in evidence, which embraces the items of ■special tax before enumerated, the non-payment of which is alleged to constitute a breach of the bond in suit, is included another special tax on 1,752¿- gallons of spirits, entered as produced in March and April, 1874, which tax amoiints to $1,226.75. This tax or assessment is not set out in the complaint as. any part of the plaintiff’s demand against O’Neill, and so it is insisted that there is a substantial and fatal variance between the allegations of the pleadings and the proofs. It is argued that this is an action of debt on the assessment; that the defendant’s answer is in effect a plea of nul tiel record; that the assessment, embracing all the items of special tax named therein, must be treated as an entirety, and as a single cause ■of action; that the items of this cause of action cannot be divided up, and separate suits maintained on each; and that since the assessment as an entirety, and as proven, does not conform in amount ■to the aggregate of the items of tax contained in the assessment described in the complaint, there is a variance fatal to the maintenance ■of the action. The answer to this is, that the suit is not, strictly speaking, upon the assessment. It is upon the bond. It is alleged ■that the conditions of the bond have been broken, in this, that the defendant O’Neill has not paid certain taxes assessed against him, and these taxes are shown in the assessment offered in evidence. In fact, the assessment only constitutes the evidence in part, of the alleged breach; and it is the breach of the condition of the bond that constitutes the cause of action. The failure to pay either of the items of tax contained in the assessment, if the tax was legally and justly imposed, would be a breach of the bond, and that would be the basis of liability. Suppose the defendant O’Neill had paid one or more of the items of tax embraced in the assessment, but had neglected to pay the other items, would not an action lie on the bond on account of such default ? Clearly it would, and so it cannot be necessary in order to maintain the action to allege and to show that there has been a default upon the entire assessment, but default may arise upon either of the items of tax, and thereupon an action for such default, based upon the conditions of the bond, may be maintained.
It is in proof that on a special assessment list of the date of November 30, 1875, there had been previously assessed against the defendant O’Neill a tax on 5,117 gallons of spirits, claimed to have ■been distilled between July 1, 1874, and March 1, 1875; that presumptively this assessment covered all the spirits manufactured and removed by the defendant during that period, and that therefore the [575] assessment in evidence, which is made the foundation of liability on the bond in suit, was unauthorized. In maintaining this contention, everything depends upon the fact whether or not the different assessments cover the same spirits. It is not shown that they do. It cannot be presumed that they do. The exorcise of authority in making the earlier assessment did not exhaust the power of the commissioner to make another assessment, embracing the whole or a part of the same period, if the two assessments did not cover the same spirits; nor does the first assessment raise such a presumption that it covered all the spirits manufactured and removed during the period named therein, as to invalidate the second and later assessment. It is, after all, a question of fact whether the two assessments cover the same spirits, and, as just remarked, it is not proven that they do.
On further review of the merits of the case, the court held that the proofs on the part of the defendant O’Neill, attacking the assessment, were not sufficient to overcome the force and effect of the assessment and the proofs adduced in its support on the part of the f government, and ordered judgment against the defendant O’Neill, and the surety, John B. Reynolds, for the sum of $25,000, the amount of the penalty of the bond.