The defendant, Charles Merrill Mount, is an art historian and former portrait painter of some note. He has been convicted under 18 U.S.C. Sec. 2314 of transporting in interstate commerce rare historical documents, knowing they were stolen from the Library of Congress and the National Archives.

On July 15, 1987, Mr. Mount wrote to Goodspeed's, an antiquarian book shop located on Beacon Hill, Boston, offering to sell a collection of letters and other documents written by the American artist, James McNeill Whistler, plus two other letters concerning Whistler, one of which was written by the French artist Henri Fantin-Latour, a Whistler contemporary. Responding to an invitation by Claire Rochefort, the director of Goodspeed's autograph department, Mr. Mount travelled to Boston from his Washington, D.C. residence on July 23, to show his collection. Ms. Rochefort agreed to buy from Mr. Mount 33 documents for a price of $20,000. Sold were 28 documents written by Whistler, two letters written by the American author Henry James, a list of articles owned by Henry James and signed by his wife Alice, a letter by Fantin-Latour, and a letter written by Winston Churchill in 1914, then First Lord of the Admiralty.

A few weeks later, on August 5, 1987, Mr. Mount again wrote to Ms. Rochefort at Goodspeed's, this time offering to sell a collection of documents from the American Civil War. In the letter Mr. Mount claimed that, while retrieving a Henry Adams letter he and Ms. Rochefort had discussed from his "vault" (actually a bank safety deposit box), "fumbling about, I came across an envelope containing a Civil War collection of 158 pieces which I thought to have been lost or stolen long ago." Mr. Mount included a list describing each document in the collection, and photocopies of the most significant ones. The collection contained eight documents written or signed by President Lincoln, plus letters (including battlefield correspondence) from Union Generals Grant, Meade, McClellan, Sherman, Pope, Hooker and Halleck, and from members of Lincoln's cabinet, Secretary of War Edwin M. Stanton, Secretary of State William H. Seward, Secretary of Treasury Salmon P. Chase, and Secretary of Navy Gideon Wells. Mr. Mount offered to sell the group of documents for $64,600.

After receiving this letter, Ms. Rochefort examined the enclosed photocopies of the Lincoln documents, and checked an eight volume set called The Collected Works of Abraham Lincoln, edited by Roy P. Basler. Basler's work contains the printed text of most of the known documents written by Abraham Lincoln, arranged chronologically. Ms. Rochefort found that the texts of three of the letters1 Mr. Mount offered for sale were contained in Basler's, and that the book credited the National Archives in Washington, D.C. as the owner of the letters. She alerted the F.B.I., and at its request invited Mr. Mount to Boston to show her the Civil War collection. Mr. Mount came to Boston the next morning, August 13, 1987. After discussing the documents with Ms. Rochefort for about 45 minutes, Mr. Mount was arrested.

Mr. Mount was indicted on two counts of violating 18 U.S.C. Sec. 2314. Count I alleged that 19 of the documents sold to Goodspeed's on July 23 belonged to the Library of Congress, that four others (the three Henry James documents and the Churchill letter) belonged to the National Archives, and that Mr. Mount knew they were stolen when he brought them from Washington, D.C. to Boston. Count II alleged that 144 of the 158 documents offered for sale on August 13 were transported from Washington, D.C. to Boston, Mr. Mount knowing they had been stolen from the National Archives.2

The government readily proved that Mr. Mount transported the documents in interstate commerce. The critical issues at trial were whether the documents had been stolen and whether Mr. Mount knew they were stolen. The government introduced evidence tending to prove that the documents had been contained in collections possessed by the Library of Congress and the National Archives and that Mr. Mount had access to the relevant collections shortly before he transported the documents to Boston. The government asked the jury to infer that Mr. Mount himself had stolen the documents, and therefore knew they were stolen.

The jury found Mr. Mount guilty on both counts. He was sentenced to three years in prison on each count, the judge suspending execution of the sentence on Count II and substituting a five-year term of probation, beginning upon release from custody.

Mr. Mount appeals this conviction. His appointed attorney (different from counsel who represented him at trial) argues (1) that the district court erred in refusing to authorize funds to secure the testimony of two foreign witnesses; (2) that the admission of certain evidence under Rule 404(b) was prejudicial error; and (3) that the trial court erred in instructing the jury. Dissatisfied with his appointed counsel, Mr. Mount requested and was granted leave to submit a supplemental brief pro se. Mr. Mount challenges (1) the sufficiency of the evidence to convict him; (2) the use of allegedly perjured testimony at various hearings; (3) the admission into evidence of several volumes of Basler's Collected Works of Abraham Lincoln; and (4) certain statements in the prosecution's closing argument.3

A. Sufficiency of the Evidence.

Evidence is legally sufficient to sustain a conviction if,

after viewing the evidence in the light most favorable to the prosecution, any rational trier of fact could have found the essential elements of the crime beyond a reasonable doubt. In making this determination, we do not assess the credibility of the witnesses, which is the sole function of the trier of fact. Nor does the government have to disprove every reasonable hypothesis of innocence, it is sufficient that the record as a whole supports a conclusion of guilt beyond a reasonable doubt.

United States v. Serrano, 870 F.2d 1, 5 (1st Cir.1989) (quoting United States v. Torres Lopez, 851 F.2d 520, 527-28 (1st Cir.1988), cert. denied, --- U.S. ----, 109 S. Ct. 1144, 103 L. Ed. 2d 204 (1989) (citations omitted). "The government may prove its case through the use of circumstantial evidence so long as the total evidence, including reasonable inferences, is sufficient to warrant a jury to conclude that the defendant is guilty beyond a reasonable doubt." United States v. Campa, 679 F.2d 1006, 1010 (1st Cir.1982).

We have extensively reviewed the record, delving beyond the parties' presentation of the facts on appeal. Theft of the documents from the Library of Congress and the National Archives was proved solely with circumstantial evidence, and was ardently contested in a spirited defense. Mr. Mount gave explanations of where and when he acquired the documents. His account was detailed and mostly consistent with the rest of his testimony. He continues to protest his innocence of any crime. The main contentions were whether the documents transported to Boston were the identical documents which had been included in the libraries' collections, and if they were, whether the libraries possessed them at the time Mr. Mount saw the collections. The defense raised some interesting questions concerning the provenance and ownership of the documents--including Whistler's purported custom of writing by hand duplicate or nearly duplicate letters, the United States' practice during the Civil War of creating various types of copies of certain communications, haphazard storage of Civil War records, and incomplete and imprecise record-keeping by the Library of Congress and National Archives. Despite this aggressive defense, we believe the evidence was sufficient to support a finding of guilt.4

1. Mr. Mount's Access to the Collections.

Shortly before approaching Goodspeed's, Mr. Mount had personally reviewed the collections in both the Library of Congress and the National Archives where the documents in question would have been kept, if they were in fact owned by the two institutions.

In 1985, the Library of Congress provided Mr. Mount a working space in its administrative office. He was not an employee of the Library. Initially, he used this space while writing an article on the American portrait painter, John Singer Sargent. During this time, he was not using the Library's reading room, where researchers gain access to the Library's manuscript collections. Instead, he worked at his desk, using his own collection of Sargent materials.

While at the Library, Mr. Mount met Randolph Boehm, who was working for a company which made microfilm sets of original manuscripts to sell to universities. The two men discussed the possibility of creating a microfilm set from the Library's Pennell-Whistler Collection, which contained some 96,000 documents written by, or concerning, the artist Whistler. (The collection is named after its donors, Joseph and Elizabeth Pennell). Mr. Boehm was interested, and beginning March 28, 1987, Mr. Mount began working in the Library's research room, making a thorough survey of the collection. His efforts culminated in a detailed report which he submitted to Mr. Boehm, claiming to have reviewed all 381 containers of materials in the Pennell-Whistler collection. Unfortunately, the microfilming project fell through when Mr. Boehm's superiors withdrew their support. Declaring that space was needed for other purposes, the Library asked Mr. Mount to relinquish his work space, which he did on June 30.

According to National Archives' records, the very next day found Mr. Mount in the Archives, asking for and gaining access to bound volumes of original manuscript records of the United States Embassy in Great Britain, records he reviewed during July and into August.5 On July 15 Mr. Mount wrote to Goodspeed's for the first time, offering to sell Whistler papers, and on July 23 he travelled there to make the sale. Returning to Washington on July 27, Mr. Mount asked to see original Civil War documents housed in the National Archives, rejecting the reference archivist's suggestion that he first consult published sources or microfilm records. According to Archives records, before writing again to Goodspeed's on August 5, Mr. Mount had been given access to all of the National Archives' holdings of letters received by the Headquarters of the Army between 1861 and 1864.

2. The Libraries' Ownership of the Documents.

Unlike the issue of access to the collections, the Library of Congress' and National Archives' prior ownership of the documents in question was hotly contested. Neither institution maintains comprehensive records of its manuscript holdings, so the government was obliged to piece together from various sources proof that they possessed these documents in the summer of 1987.

Two defense theories were raised. First, because various duplicates of some documents allegedly exist, Mr. Mount argued that both he and the libraries could have owned original documents, the libraries' records of ownership actually proving that they owned documents different from the ones Mr. Mount transported to Boston. Second, even if the documents were the same, the proof in many instances was that the documents were in the possession of the libraries long enough in the past so that they could have been removed from the collections, later innocently to turn up in Mr. Mount's hands. The government conceded that at one time, at least, Mr. Mount collected original manuscripts of persons in whom he had a scholarly interest.

(a) The Library of Congress. We briefly set forth the evidence showing that the 19 Whistler documents listed in Count I were part of the Library of Congress' Whistler collection.

While working on the Pennell-Whistler microfilm project proposal in the Library's reading room, Mr. Mount often showed or read to Mr. Boehm letters he found especially interesting. At trial Mr. Boehm testified that he remembered one of these letters in particular--Whistler writing to Frederick Leyland that he had finished painting the walls of the elegant "Peacock Room" in Leyland's London house. (The walls of this room were later removed, and now reside in the Freer Gallery in Washington, D.C.) The Library of Congress' reading room rules prohibit patrons from bringing in any materials other than special note paper, so it is a fair inference that the letter was part of the Pennell-Whistler collection. Mr. Mount claimed at trial that the letter was part of his own collection, and that he read it to Mr. Boehm in his work space in the administrative office, but the jury was free to believe Mr. Boehm's version.

Five Whistler letters identical to ones sold by Mr. Mount (including The Peacock Room letter) appear in a partial microfilm catalogue of the Pennell-Whistler Collection made by the Library of Congress in 1963. According to Mr. Mount, and there is some evidence supporting him, Whistler had the habit of writing by hand identical or nearly identical letters. However, the jury saw both the letters Mr. Mount sold, and the photocopies made from the Library's microfilm catalogue, and could conclude based on comparison that the letters Mr. Mount sold were the same ones the Library owned in 1963.

Nine other Whistler letters are listed in acquisition records which contain short descriptions or quotations from manuscripts acquired by the Library between 1968 and 1979. The acquisition record descriptions are not equally detailed, and three vary slightly when compared to the text of the original letters. However, the record for one letter acquired in 1979 includes a photocopy of the back of the document, and envelopes corresponding to the letters remain in the Pennell-Whistler collection, while the letters are missing.6 The jury could conclude that the variances between the letters and the recorded descriptions were merely clerical mistakes.

Finally, Kenneth Rendell, a private manuscript dealer identified from his invoice records eight Whistler letters which Mr. Mount had sold to Goodspeed's (including three that also appear in the Library's acquisition records) as letters he had sold to the Library of Congress between 1972 and 1975. The dealer recognized on the manuscripts his own pencilled catalog numbers.

(b) The National Archives. The government offered some proof that the National Archives owned each of the 148 documents alleged to belong to it in Counts I and II. Although the proof is stronger for some documents than others, it is sufficient to support a finding of guilt.

The government showed that four of the documents sold on July 23 (the Henry James and Winston Churchill documents) at one time had been contained in three bound volumes of the Archives' United States Embassy collection--records which Mr. Mount saw just prior to the sale.7 The four documents match, both physically and by content, places in the volumes where documents are missing.

As for the Civil War materials, National Archives Reference Archivist Michael Musick testified that Mr. Mount described to him a document which Mr. Mount found especially interesting--a letter from Union General John Pope to General Halleck, in which Pope had "justified his entire military career." (Pope, after being blamed for the Army of the Potomac's defeat at the Second Battle of Bull Run, had been assigned to a remote command in Minnesota.) At trial, Mr. Musick identified one of the manuscripts which Mr. Mount had tried to sell to Goodspeed's on August 13 as the letter Mr. Mount had described to him in the Archives' reading room. Like the Library of Congress, the National Archives allows researchers to bring no material except special note paper into its reading room. If Mr. Musick was correct, it can be inferred that the National Archives owned the letter.

Sara Jackson, who worked for the National Historical Publications and Records Commission, produced photocopies of two Lincoln documents identical, except for certain minor differences (see below), to those offered by Mr. Mount. She testified that the photocopies were made from original documents in the National Archives collections, one in 1981, and the other sometime after she joined the Commission in 1968 or 1969.

Colonel Arthur V. Grant testified that while writing a Master's thesis in 1973 on the command relationship between Generals Grant and Meade, he viewed the National Archives collection of letters received by the Army Headquarters between 1861-1864. Colonel Grant produced photocopies of six original documents he had made from that collection, each identical to ones Mr. Mount had offered to sell to Goodspeed's. The government also introduced 44 photocopies made from a microfilm edition of the correspondence of Ulysses S. Grant identical to documents in the group Mr. Mount offered to Goodspeed's. This microfilm project was conducted from 1965 into the late 1970's.

Responding to Mr. Mount's "multiple copy" theory, Mr. Musick, qualified to testify as an expert, described the copying techniques available during the Civil War. According to Mr. Musick, official correspondence could appear in "fair copies," the most common, which were longhand copies written out by clerks hired for their clear penmanship; "press copies," which were made by pressing a very thin sheet of paper onto the original while the ink was wet; "telegraph copies," created by the receiving telegraph operator as he wrote down the incoming message; "manifold copies," described as a precursor of carbon paper; "photographic copies," made by actually taking a photograph of the original; and "author copies," where the author actually wrote out in his own hand, two copies of the document.

Mr. Musick testified that documents created by each of these copying techniques (except author copies) are readily identifiable as copies. He claimed that none of the Civil War documents possessed by Mr. Mount is such a copy, and that none of the National Archives microfilm was made from such copies. He also testified that he was not familiar with commanders making author copies of battlefield correspondence, which makes up a large portion of the Civil War documents in question. John Simon, the head of the Grant microfilming project, also testified that the microfilm record was made from original manuscripts, not from documents created by then-current copying techniques. It also seems impossible that the two Lincoln endorsements credited to the National Archives in Basler's Collected Works of Abraham Lincoln, written as they were on the back of incoming correspondence, could have been copies. The jury could properly decide that Mr. Mount and the National Archives possessed the same documents, not two sets of copies.

Mr. Musick rounded out the government's proof of ownership by giving his opinion that he was "98-99% certain," after consulting "letters registers" (contemporaneous records kept of letters received by the Union Army headquarters during the Civil War) held by the National Archives, that the remainder of the documents in question should have been contained in the Archives collection in 1987. He testified specifically that ten of these documents were missing from the collection, based on his examination of related documents still in the Archives.

We recognize that except for the letters Mr. Mount read to Mr. Boehm and Mr. Musick, the government's proof of the presence of the documents within the two collections predates, often by many years, Mr. Mount's access to the collections. In order to find guilt, then, the jury had to infer that the documents stayed put from the time of proof of ownership until the time Mr. Mount examined the collections in 1987. Although Mr. Mount's "innocent acquisition" defense is not wholly implausible (there was evidence that stolen documents were sold in the manuscript market), as we demonstrate below, Mr. Mount's explanations of where and when he acquired the documents were largely inconsistent with possession by the libraries at the times proved by the government. The jury was required to believe either the government's evidence of ownership, or Mr. Mount's explanations.

According to Mr. Mount's testimony, he acquired the Whistler letters as a private collector during the 1950's and 1960's, but stopped collecting original manuscripts in 1969. Much of the Library of Congress' proof of ownership of documents postdates 1969--the reading of the Peacock Room letter to Mr. Boehm, several of the Library's acquisition records, and Mr. Rendell's invoices covering eight of the documents--so that if the Library did possess these documents after 1969, Mr. Mount could not have acquired them as he said he did. The jury apparently credited the Library's proof of ownership, and disbelieved Mr. Mount. Similarly, Mr. Mount claimed he got the Civil War documents sometime between 1961 and 1963 from Father Grogan, "a garrulous Irish priest," but much of the proof of the National Archives' ownership of these documents postdates 1963--Mr. Mount's description to Mr. Musick of the General Pope letter, the making of the two photocopies identified by Ms. Jackson, and the 44 General Grant photocopies and microfilm. Again, to credit the Archives with ownership of the documents is to reject Mr. Mount's defense.

It is theoretically possible, even accepting the government's proof of ownership and rejecting Mr. Mount's copies defense, that some documents still could have ended up in Mr. Mount's hands innocently. For example, the four documents from the Embassy volumes could have been removed prior to 1969 (the volumes were already bound when received by the National Archives in 1949 and 1950), Mr. Mount later acquiring them innocently. It is also theoretically possible that Mr. Mount innocently acquired the three Whistler documents which the Library of Congress proved it owned only by acquisition records dated before 1969, and quite a few of the Civil War documents. Yet, in light of the government's proof of ownership of the other documents and its inconsistency with Mr. Mount's defenses, these possible explanations are not compelling. The jury could fairly infer that these documents, too, were stolen.8

The district court denied motions by Mr. Mount asking the government to provide funds to pay the travel expenses of three witnesses (one from Iowa and two from Ireland), each of whom he argues on appeal had seen him in possession during the 1960's of documents similar to the ones in question.

The Iowa witness, Barbara Thompson, eventually paid her own expenses and testified at trial. She testified to becoming acquainted with Mr. Mount in Ireland in 1961, at which time he showed her a collection of old documents. At trial she noted many similarities between the documents she saw in Ireland and those which Mr. Mount offered to sell to Goodspeed's, such as the color, size, style of handwriting, and content--that they contained "military terminology" and referred to generals, troop movements, Grant, Lincoln, and Appomattox. However, she could not identify any specific document, only saying that the ones taken from Mr. Mount at the time of his arrest "appear to be similar" to those she saw in Ireland in 1962. After she testified, the court denied a renewed motion for funds for the other two witnesses.

In an ex parte affidavit submitted to the district court, Mr. Mount described what he believed the other two witnesses, citizens and residents of Ireland, could have testified to. According to Mr. Mount, James J. Magill, Mr. Mount's doctor when he lived in Ireland, could have testified to seeing Civil War documents in the 1960's at Mr. Mount's Irish residence. He purportedly recalls the name "General Sherman," and remembers a glass-topped display table in which Mr. Mount displayed original manuscripts. The second, Thomas Ryan, again according to the defendant, could have testified to seeing Mr. Mount in possession of manuscript letters by John Singer Sargent, Whistler, and Joseph Pennell, among others. He also purportedly recalls the display table and filing cabinets in which Mr. Mount kept his manuscript collection. The affidavit is silent as to whether either witness could recall any specific document. At trial and on appeal, Mr. Mount's counsel conceded that no showing was made that either could do so. Mr. Mount claims that he was entitled to have the government pay Dr. Magill's and Mr. Ryan's travel expenses under Criminal Rule of Procedure 17(b), his sixth amendment right to compulsory process, and his fifth amendment right to due process of law.

Rule 17(b) does not give the district court authority to order payment of expenses of witnesses outside the subpoena power of the court. Although the Note to the 1966 amendment to Rule 17(b) says that "an indigent defendant will be able to secure the attendance of witnesses at the expense of the Government no matter where they are located," the power to order payment of witness fees and costs under the Rule is coextensive with the power to subpoena.

Rule 17 ... is limited to proposed witnesses whom the court has power to summon and for whose travel expenses there is statutory authority. A person who at the time of the Rule 17 application is not a citizen or resident of the United States and is not within the United States does not come within that limited category.

United States v. Gordon, 634 F.2d 639, 645-46 (1st Cir.1980).

Nor do we think the court's refusal to provide travel expenses violated the Constitution. Before the absence of defense witnesses can be said to violate either the right to compulsory process or due process, the defendant must show that the testimony from the missing witnesses would have been relevant, material, and favorable. United States v. Valenzuela-Bernal, 458 U.S. 858, 867, 872, 102 S. Ct. 3440, 3446, 3449, 73 L. Ed. 2d 1193 (1982); United States v. Hoffman, 832 F.2d 1299, 1303 (1st Cir.1987) (compulsory process); United States v. Bartlett, 794 F.2d 1285, 1291 (8th Cir.), cert. denied, 479 U.S. 934, 107 S. Ct. 409, 93 L. Ed. 2d 361 (1986) (due process).

While there is not much question that the Irish witnesses' testimony that Mr. Mount possessed in the 1960's documents similar to those he transported to Boston in 1987 would have been minimally relevant to his defense, Mr. Mount's affidavit does not show that Dr. Magill and Mr. Ryan could have added anything of substance to Ms. Thompson's testimony. That Mr. Mount possessed a display table and filing cabinet does not prove his ownership of specific documents; at best it corroborates a tangential element of his story. Both Ms. Thompson and Mr. Mount testified at trial that Mr. Mount possessed Civil War documents when he lived in Ireland, and we conclude he was not denied due process by the trial judge's refusal to order payment of expenses of the Irish witnesses. See, e.g., Bartlett, 794 F.2d at 1291 (no prejudice shown from death of potential witnesses who could only testify to what others at trial did); United States v. Pino, 708 F.2d 523, 528 (10th Cir.1983) (no prejudice from death of witness without showing that the witness would have testified differently from one who did).

Mr. Ryan apparently could have added one point which Ms. Thompson did not testify to--that Mr. Mount had one or more manuscript letters by Whistler when he lived in Ireland. There is no indication that he could have gone beyond this degree of generality. The government did not dispute that Mr. Mount, a biographer and art historian, at one time owned original manuscripts; at issue was his ownership of the eighteen Whistler letters listed in Count I of the indictment. Before asking the government to bear the cost of bringing over one witness who could say no more than that Mr. Mount previously possessed one or more unidentified Whistler letters, it was incumbent upon the defense better to explore the extent of Mr. Ryan's memory.10 The court did not err in refusing to fund Mr. Ryan's travel expenses. See Hoffman, 832 F.2d at 1305; Bartlett, 794 F.2d at 1291. Cf. United States v. Mateos-Sanchez, 864 F.2d 232, 239-40 (1st Cir.1988) (travel expenses for investigatory trip to Spain properly denied under Criminal Justice Act where defendant failed sufficiently to explain reason for trip).

This case is obviously quite different from United States v. Largan, 330 F. Supp. 296 (S.D.N.Y.1971), the only case cited by defendants where witness travel expenses were ordered paid by the government. There, the defendant was a Philippine resident and citizen brought to the United States for trial. Finding the case to be "exceptional" in that "all of the defendant's prospective witnesses may be unable to attend or prevented from attending his trial," and that the missing witnesses' testimony was material, the court ordered payment of expenses for the defense counsel to travel to the Philippines to take depositions, as well as travel expenses for two character witnesses to attend trial. Id. at 298. In contrast, the testimony of Mr. Mount's two Irish witnesses, as reflected in his own affidavit, was mostly cumulative and not directly probative of the question at issue.

C. Evidence of Removal of Identifying Marks.

F.B.I. Special Agent Perrotta testified at trial that he had examined a number of the documents in question, and concluded that fourteen of them showed evidence of having been tampered with--specifically, that identifying marks of some sort had been removed.

Also, another F.B.I. agent testified that after Mr. Mount was arrested, his Washington, D.C. boarding house room was searched, and fourteen other documents (not the subject of this prosecution) were found, pressed within the pages of a large book. Each document was located between blank sheets of paper, or between blank pages of the book. Two of these, letters written in 1920, were admitted into evidence. Agent Perrotta testified that identifying marks also had been removed from these. One letter previously had a stamp which read "London, American Embassy" and the letters "A.B.L.", the other, the initials "A.B.L." Agent Perrotta gave his opinion that most of the obliterations he discovered had been done using a liquid bleach.

The defendant argues that this evidence was inadmissible under Federal Rule of Evidence 404(b).11 We conclude there was no abuse of discretion.

First, assuming that the evidence of obliteration of identifying marks from the fourteen documents included in the indictment somehow is evidence of "other acts" subject to 404(b), it went to prove something other than Mr. Mount's bad character. For a number of the charged documents, the photocopies or microfilm introduced by the government to prove ownership showed certain identifying stamps or marks on the documents which those taken from Mr. Mount did not. Agent Perrotta's testimony that certain documents taken from Mr. Mount had marks removed offered an explanation for these differences.

Also, this evidence helped show that Mr. Mount knew the documents were stolen. Removing identifying marks is consistent with a desire to conceal the origin of stolen material. While it is true that the government did not prove when the obliterations were done, the defendant does not dispute that he had the opportunity to perform the alterations. We think the evidence was sufficient to support a conclusion that Mr. Mount had performed the obliterations, and so was admissible under 404(b). See Huddleston v. United States, 485 U.S. 681, 685, 108 S. Ct. 1496, 1499, 99 L. Ed. 2d 771 (1988). Mr. Mount's contention that some previous holder of the documents could have made the changes merely disputes the probative weight of the evidence, not its relevance.12

The relevance of the two documents found in Mr. Mount's room, and the testimony concerning obliterations found on them, is less readily apparent. Since the government did not try to prove that the National Archives or Library of Congress owned these documents, there was no need to explain variations between the government's proof of ownership and the documents possessed by Mr. Mount. However, we think the evidence helped marginally to connect Mr. Mount with the obliterations on the charged documents (by showing that other documents in his possession had similar obliterations) and with the obliteration technique hypothesized by Agent Perrotta (the documents may have been pressed between pages in the large book in order to absorb a bleaching agent), and hence help show participation in and knowledge of the theft of those charged documents showing obliterations. It also helped show that the obliterations on the documents possessed by Mr. Mount may have been recently performed, shoring up the government's theory that Mr. Mount, not some prior possessor of the documents, had performed the obliterations.

Nor was there clear error or abuse of discretion in the district court's conclusion that any unfair prejudice from this evidence did not substantially outweigh its probative value under Rule 403. See United States v. Garcia-Rosa, 876 F.2d 209, 220 (1st Cir.1989). This court has stated that evidence should be excluded for unfair prejudice "only where the trial judge believes that there is a genuine risk that the emotions of the jury will be excited to irrational behavior, and this risk is disproportionate to the probative value of the offered evidence." United States v. Fields, 871 F.2d 188, 198 (1st Cir.), cert. denied, --- U.S. ----, 110 S. Ct. 369, 107 L. Ed. 2d 355 (1989) (citations omitted). We do not think the trial judge erred by deciding that this evidence would not improperly inflame the jury.

D. Jury Instructions.

Mr. Mount argues that the court's instructions "diluted" the standard of proof beyond a reasonable doubt.

The argument is directed at the court's statements (1) that "the government's theory, nevertheless, is that [defendant] knew [the documents] were stolen because he was the one that stole them" and (2) several sentences later, "If you find that the Government has shown that he stole them, you may infer from that that he knew they were stolen."

The argument appears to be that without modifying the word "find" in (2) with "beyond a reasonable doubt," the jury may have been led to believe that it could find Mr. Mount had knowledge of the stolen character of the documents without being convinced beyond a reasonable doubt that Mr. Mount had stolen them.

This argument cannot succeed unless the omission was plain error affecting a substantial right. Fed. Rules of Crim. Proc. 30 and 52. See United States v. Potes, 880 F.2d 1475, 1478 n. 1 (1st Cir.1989). Counsel made no objection to the omission, but suggested only that sentence (1), quoted above, was an "impermissible comment" and an endorsement of the government theory.

This happens to be a case where much of the evidence tending to show that the documents had been stolen, and the only evidence that Mr. Mount knew of their stolen character, was evidence tending to show that he stole them. An instruction telling the jury that it could find Mr. Mount guilty without being satisfied beyond a reasonable doubt that he stole the documents would, under these facts, have been erroneous. The instructions as a whole, however, made crystal clear that the jury could not convict unless it found all the essential facts beyond a reasonable doubt.

At one point during the instructions, Judge Zobel said: "If there remains in your mind a reasonable doubt as to the existence of any fact that is necessary to prove either Count 1 or Count 2, then, the defendant is entitled to the benefit of that doubt and you cannot find him guilty." The court's instructions contained four other statements that guilt, or the elements of the offense, must be proved beyond a reasonable doubt.

Judge Zobel listed the elements of the offense, one being that at the time the documents were transported, they had been stolen from their owner, and another being that the defendant knew they were stolen. The sentences now complained of were part of the discussion of the latter element. The jury was clearly told that that element, like the others, must be proved beyond a reasonable doubt, and it would require a strained interpretation of the language to think that a finding that Mr. Mount had stolen the documents, arrived at without being convinced beyond a reasonable doubt, could be the predicate for a finding that he knew they had been stolen.

We are convinced that the omission of an additional "beyond a reasonable doubt" phrase in the sentence complained of had no significant effect on the verdict, and was not plain error affecting a substantial right, if it was error at all. United States v. Jarabek, 726 F.2d 889, 904 (1st Cir.1984).

E. Use of "Perjured" Testimony.

The defendant also claims that an FBI special agent committed perjury at Mr. Mount's detention hearing, before the grand jury, and at trial. At the detention hearing, the agent concluded that three of the documents Mr. Mount sold to Goodspeed's belonged to the National Archives, after comparing them to documents appearing in what he called "a book of National Archive museum paraphernalia." Transcript of August 14, 1987 Hearing before Magistrate Alexander at 36.13 Mr. Mount characterizes this as an assertion that the documents appear in a book published by the National Archives. In truth, the texts of the documents appear in Basler's Collected Works of Abraham Lincoln, which is published by Rutgers University Press.

This misattribution is not perjury, unless the agent did not believe what he said to be true. See 18 U.S.C. Sec. 1621. At trial, the agent explained that he had forgotten the name of the book, and was trying to explain his understanding of the contents of the book. Transcript of Trial, Vol. II at 151. We see no reason to disbelieve this explanation, nor how the misattribution could make any difference anyway. There is no dispute that The Collected Works of Abraham Lincoln actually contains the text of the three Lincoln documents and credits them to the Archives. The agent's testimony showed, correctly, that an authoritative source says those three documents belong to the Archives, not Mr. Mount.

Even if the minor inconsistency was material, it was harmless. See United States v. Mechanik, 475 U.S. 66, 70, 106 S. Ct. 938, 941, 89 L. Ed. 2d 50 (1986) (petit jury's finding of guilt makes procedural error in grand jury proceeding harmless). Witness perjury in establishing probable cause becomes harmless after the rendition of a guilty verdict unless prosecutorial misconduct is so "serious and blatant" as to "distort[ ] the integrity of the judicial process," United States v. Bucci, 839 F.2d 825, 831 (1st Cir.), cert. denied, --- U.S. ----, 109 S. Ct. 117, 102 L. Ed. 2d 91 (1988) (quoting United States v. Ogden, 703 F.2d 629, 636 (1st Cir.1983). A fortiori, perjury at a detention hearing would not taint a properly supported verdict of conviction.

In successive submissions, Mr. Mount has also pointed out that the same agent, when testifying before the grand jury, gave what Mr. Mount considers to be false testimony concerning the National Archives' ownership of one of the documents included in Count II. We agree that the agent's testimony was inaccurate in certain details (as to when the Archives first gained possession of Civil War records and how accurately a letters register describes the document in question), and may have given an inflated impression of the strength of the evidence of the Archives' ownership of that one document. However, the agent also presented to the grand jury the Archives' summary of its documentation of ownership of the other 143 Civil War documents, and this was enough proof of ownership to support the indictment.

F. Admission of Basler's.

The two volumes of The Collected Works of Abraham Lincoln which Ms. Rochefort of Goodspeed's consulted after receiving Mr. Mount's second letter were admitted into evidence. Mr. Mount claims the volumes were inadmissible, because under Federal Rule of Evidence 803(18) learned treatises cannot be admitted as substantive evidence except to "the extent called to the attention of an expert witness upon cross-examination or relied upon by the expert witness in direct examination," and even then only to be read into evidence, not received as exhibits. Fed. Rule Evid. 803(18) & Note. Regardless of their admissibility under Rule 803(18), the volumes comfortably fit within the hearsay exception in Rule 803(17), which allows admission of "tabulations, lists, directories, or other published compilations, generally used and relied upon by the public or by persons in particular occupations." Ms. Rochefort testified that manuscript dealers like herself rely on Basler's work to locate original Lincoln documents, so the foundational requirements of Rule 803(17) were met.

Mr. Mount also claims the volumes are unreliable evidence because of their age (they were published in 1953), so they "shed no pertinent light on recent ownership of any document." The volumes tended to prove possession of the three documents by the National Archives in 1953, and by inference, in 1987, so they were relevant. It is possible, as Mr. Mount argues, that the three Lincoln documents may have left the National Archives after 1953; however, Judge Zobel was within her discretion to let the jury evaluate the weight of the volumes as proof of the Archives' ownership in 1987.

G. Closing Argument.

During his closing argument, the prosecutor asked the jury to find that Mr. Mount had stolen the documents, on its way to concluding that he knew the documents were stolen when he took them to Boston. Although Mr. Mount's brief does not specify which comments he finds objectionable, he apparently means the following statements:

Did Mr. Mount know that they were stolen when he brought them to Boston? I suggest to you that the answer, again, is clearly yes. How did he know that they were stolen? The Government has proceeded from the beginning on a simple theory, on the theory that he knew that they were stolen because he was indeed the one who stole them.

The prosecutor then asked "How do we know that he stole the documents from the Library of Congress?" and went on to review the evidence.14 Mr. Mount claims this was improper and prejudicial because "[t]hrough ten days of trial no evidence, no witness, and no assertion concerning a theft blemished the record," and so the argument had no basis in the record. Mr. Mount's trial counsel did not object to any portion of the prosecution's argument, nor request a curative or limiting instruction. Therefore, we review only for plain error. United States v. Machor, 879 F.2d 945, 955 (1st Cir.1989); Fed. Rule of Crim.Proc. 52(b).

Although it is the jury's job to draw inferences, there is nothing improper in the Government's suggesting which inferences should be drawn. United States v. Cotter, 425 F.2d 450, 452 (1st Cir.1970). See United States v. Doherty, 867 F.2d 47, 70-71 (1st Cir.), cert. denied as Deliere v. United States, --- U.S. ----, 109 S. Ct. 3243, 106 L. Ed. 2d 590 (1989) (evidence supported arguing conclusion that defense witnesses were "friends" of the defendant). Such argument is improper only when the prosecutor's own credibility is inserted into the argument. United States v. Rosa, 705 F.2d 1375, 1379 (1st Cir.1983); Cotter, 425 F.2d at 453.

Because, as we have held, there was enough evidence to support an inference that Mr. Mount had taken the documents, the prosecutor was entitled to ask the jury to reach that conclusion. The prosecutor did argue that based on the evidence, the government's theory was the more persuasive, but that is the whole point of argument. Unlike the prosecutor in United States v. Cresta, 825 F.2d 538, 555-56 (1st Cir.1987), cert. denied as Impemba v. United States, 486 U.S. 1042, 108 S. Ct. 2033, 100 L. Ed. 2d 618 (1988), who said "I think it would be ridiculous to draw any type of inference like that," the prosecutor here avoided expressing any personal conclusions.

Nor did the prosecutor's question "How do we know that he stole the documents from the Library of Congress?" imply that he knew of additional evidence of guilt not presented to the jury. Read in context, this is obviously just a rhetorical technique, as he immediately followed the question by recounting the evidence introduced at trial supporting the government's case.

For the foregoing reasons, the conviction is AFFIRMED. The several motions filed by Mr. Mount in this court and not heretofore disposed of are DENIED.