The meeting minutes were the first document we had seen that recorded a direct decision by the board rather than an instruction from the legal office or an entry in an accounting ledger. Their wording was unambiguous. The directors had agreed to delay the release until the claimant schedule and supporting payment records were complete. No exception had been approved for repayment of the private advance, and no authority had been granted to treat unresolved claims as settled.
Leonard Voss had signed the minutes as secretary.
Miriam explained that the document needed to be authenticated and compared with the surviving copies held by other directors. A secretary’s signature did not automatically prove that every statement in the minutes was accurate, but it did make Voss’s claim of a board-approved exception difficult to reconcile with the official record.
Rebecca read the final paragraph again. “If he recorded the decision himself, he knew the release was supposed to wait.”
“Unless he later received a separate instruction from the board,” Miriam said. “We have to check whether another meeting occurred or whether the directors issued a written resolution. So far, the archive index shows no such document.”
The investigators contacted Harold Finch and two other surviving directors. Finch confirmed that the meeting had taken place and that the minutes reflected the decision as he remembered it. He said no later vote had authorized the private repayment. The other directors had differing memories of the exact discussion, but neither recalled approving an exception that allowed the reserve to be transferred before claimant payments were verified.
One director remembered Voss saying that the legal office could prepare the release once the remaining administrative issues were resolved. He did not remember being told that a private investor would be repaid from the reserve before those issues were completed. Another said he had assumed the legal department would follow the board’s decision and did not question the subsequent accounting summary because it showed the settlement as finalized.
Their recollections were not identical, but they supported the minutes on the essential point: the board had not knowingly authorized the disputed sequence of transfers.
The investigators then compared the minutes with the second release. The release referred to the board’s approval as the basis for finalizing the settlement, but it did not cite a meeting date or resolution number. The private agreement, meanwhile, included a clause allowing the transfer to proceed if Arthur declined approval, provided that the legal office certified the transaction.
That clause appeared to have been used as a substitute for the authority the board had withheld. The agreement did not expressly say that Voss could override the board, but its structure allowed the legal office to proceed while keeping the private repayment outside the ordinary claimant ledger.
Miriam cautioned us against calling the arrangement a deliberate conspiracy before the evidence was complete. The records could support several explanations, including a knowingly unauthorized transaction or a misguided attempt to satisfy an investor while promising to reconcile the claims afterward. The difference would matter legally, even though both explanations raised serious questions about the missing payments.
Still, the documents were becoming increasingly difficult to interpret as a simple mistake. Arthur had objected in writing. Patricia had preserved a schedule showing unresolved claims. Daniel had refused to certify the altered reconciliation. The board minutes required the release to wait. Yet the transfer went forward, the register was changed, and the private agreement was used to repay an investor whose identity had been concealed from the ordinary records.
The next discovery came from the trust administrator. The separate reconciliation reserve had not been spent immediately. It remained in a holding account for several months after the transfer, then was moved into a trust that required Arthur’s confirmation before the money could be released. The trust had been created because the account administrator could not determine whether the reserve belonged to the company, the investor, or the unresolved claimants.
That uncertainty was important. Even the people managing the money had not been given a consistent explanation of its ownership. The main transfer treated the settlement as complete, while the separate reserve remained subject to a condition that could only be satisfied once Arthur verified the claimant schedule.
The trust correspondence included a letter from Arthur dated January 1988. He wrote that he would not provide the requested confirmation because the three disputed claims remained unresolved and the supporting receipts had not been produced. He asked the administrator to freeze the reserve until the original schedule and all payment records could be examined.
He also stated that the second release should not be treated as proof of final settlement because he had not signed it and had not authorized anyone to sign on his behalf.
The letter was direct and specific. It did not accuse any individual of forging his signature, but it made clear that he rejected the release as evidence of his approval. The date placed it after the disputed transfer but before the reserve was moved into the trust.
Miriam said the letter gave the investigators a contemporaneous record of Arthur’s position. It also created a problem for anyone claiming that the release had been properly authorized. If Arthur had expressly rejected the document and refused to confirm the reconciliation, the legal office could not reasonably rely on his approval without explaining how it had been obtained.
The trust administrator had responded by refusing to release the remaining funds. The correspondence showed that Voss later asked for the restriction to be lifted, arguing that the final release had resolved the settlement. The administrator declined, citing Arthur’s objection and the absence of a complete claimant schedule.
A second letter from Voss followed several weeks later. It asked the administrator to reclassify the reserve as a legal hold rather than a claimant reserve. The request did not succeed, but it showed that Voss had attempted to change how the money was categorized after Arthur refused to confirm the reconciliation.
That detail matched the pattern in Martin Keene’s notes: documents and accounts were repeatedly moved into different categories when the original record became inconvenient. The changes did not make the underlying obligations disappear. They made the obligations harder to trace.
Rebecca added the trust correspondence to her spreadsheet. The sequence now extended beyond the initial transfer. Arthur had rejected the release in January 1988, and the trust administrator had kept the remaining reserve frozen. Voss had later tried to have the funds reclassified, but the restriction remained in place.
“What happened to the reserve after that?” I asked.
Miriam said the records showed a further transfer several years later, but the destination had not yet been identified. The trust had eventually been closed after a legal review, and the remaining funds had been moved to an account administered by a successor firm. The closure documents cited a determination made under a separate reference number.
The reference number was familiar.
It matched the one on the envelope marked Original signature page—retain separately.
The connection suggested that the private agreement and the trust reserve had been handled together at some point, even though they were supposed to represent different parts of the transaction. One account had carried the principal repayment. The other had held funds pending proof that the claimant schedule was complete. The same reference system had been used to keep both arrangements outside the ordinary register.
The investigators requested the closure file from the successor firm. It contained a short memorandum stating that the reserve could be released once the legal office confirmed that the final settlement had been properly authorized. The memorandum was dated after Arthur’s objection and was signed by a person acting for Voss’s office.
There was no evidence that the trust administrator had received the board minutes or Patricia’s original schedule. Instead, the memorandum relied on the same second release that Arthur had rejected.
Miriam said the closure file might explain why the reserve had eventually been moved despite the unresolved claims. But the investigators still needed to identify who received the funds and whether any part of the money had later been returned to the company or distributed to the claimants.
They also needed to establish whether the signature on the second release was forged, copied, or otherwise used without authorization. The examiner had already identified discrepancies between the signature and authenticated examples of Arthur’s handwriting. The board minutes and Arthur’s written objection strengthened the case that he had not approved the transaction, but the signature itself remained a separate evidentiary question.
A week later, Miriam called with the examiner’s preliminary findings. The disputed signature contained features inconsistent with Arthur’s authenticated writing, including differences in the formation of two letters and the pressure pattern through the final stroke. The examiner could not yet determine whether the signature had been traced, copied from another document, or produced by someone attempting to imitate Arthur’s hand.
The examiner also found that the signature appeared to have been placed on the document after the main text had been printed. That did not establish the exact sequence of preparation, but it raised further questions about how the release had been assembled.
The investigators asked Voss to provide the original file from which the release had been prepared. His representative stated that the original document could no longer be located. The legal office had retained a copy, but the underlying paper had been transferred during an office move and was presumed lost.
Miriam said that explanation would need to be checked against the archive records. Martin Keene’s notes indicated that an original authorization had been moved into a restricted envelope and assigned a different reference. If the original paper still existed, it might be among the documents already recovered.
The investigators searched the restricted archive again and located an envelope bearing the same reference as the private agreement. Inside was a document described as an authorization for the final release. It carried Arthur’s name, but the signature block was blank. A handwritten note at the top said: Authorization not completed. Await confirmation.
The document appeared to be an earlier version of the release, prepared before the signature was added to the copy that later entered the official register. It showed that the legal office had possessed a version of the authorization without Arthur’s signature. The note also indicated that someone knew confirmation was still required.
Miriam asked us not to treat the blank signature block as proof that the final signature was forged. The document might have been a draft that was later signed properly. But when compared with Arthur’s written objection and the examiner’s findings, it became another reason to question the release’s authenticity.
At the bottom of the page, beneath the note about awaiting confirmation, was a short instruction in a different hand: Proceed under legal certification if confirmation is not received.
The instruction was dated December 11, 1987—the day the reserve was transferred.
The wording matched the clause in the private agreement that allowed the transaction to proceed if Arthur declined approval. It appeared that the alternative had been prepared before anyone had obtained the confirmation the board required.
The investigators now had evidence that the release had been processed under a legal-certification procedure despite the missing authorization. They still needed to identify who invoked that procedure, whether it was permissible under the company’s rules, and how the disputed signature came to appear on the final document.
Miriam asked us to meet her the next morning. She had received one more item from the archive: a covering letter attached to the blank authorization. It listed the documents that should have been completed before the release could be finalized. Among them were the original claimant schedule, the payment confirmations for Clara Whitcomb, the Miller family, and Bell Equipment, and Arthur’s written approval.
Beside each item was a check box.
All three claimant confirmations were marked incomplete.
Arthur’s approval was marked incomplete as well.
Yet the final release had been processed the same day.
I stared at the four empty boxes and thought of the years our family had spent trying to understand why Arthur had been so certain that something was wrong. He had not merely suspected a discrepancy. He had been told that the documents were incomplete, had written his objection, and had refused to confirm the reconciliation. The transfer went forward anyway, and the official record was made to say the opposite.
The final question was no longer whether the board had approved the release. The minutes showed that it had not. The question was who had decided that legal certification could replace the missing approvals, who had placed Arthur’s signature on the final copy, and whether that person had acted to protect the private beneficiary or to protect someone else from being exposed.
Miriam closed the folder.
“We are getting closer,” she said. “But the next step matters more than any conclusion we can draw today. We need the person who gave the instruction to explain it under oath—or the original record that shows exactly who did.”
Click here to continue reading: PART 29: The Missing Authorization Finally Reveals Who Ordered the Release and Why Arthur Brooks Refused to Stay Silent
My Mother Wrote Six Ordinary Numbers, but the Man Behind the Counter Suddenly Forgot How to Breathe
Part 28 of 30
