Helen Brooks began examining Mark’s recovered files the following morning. She worked from copies, leaving the original devices and storage media untouched until Martin could arrange a proper technical review. The collection contained several versions of the transaction summary, but the files had been transferred between systems, and some of their original timestamps were missing. Helen warned Eleanor that a document’s name could suggest a sequence without proving it. A file labeled “final” might have been saved before a file labeled “draft,” and an old copy could have been renamed long after the underlying text was written.
The first comparison confirmed that Mark had not simply copied one complete summary into another. The earliest recovered version described the reconciliation certificate as a document received during the closing review. It said that the supporting schedule remained under examination. A later version removed that qualification and stated that the final accounting had been approved by Peter and David. The wording in the later version was nearly identical to the sentence in the summary Mark had ultimately circulated.
The sequence was troubling, but the recovered files did not establish who had made the change. One version appeared to have been edited through a shared project account, while another had been saved under Mark’s account. Because the files had been copied between computers, the account information did not reliably identify the person who had typed each sentence. Helen could determine which text had changed, but not who had made every change or what instructions had accompanied it.
Eleanor read the comparison at Martin’s office. Claire sat beside her, tracing the changes with the tip of a capped pen. The repeated phrase about approval seemed almost mechanical, as though the final account had been shaped to remove every uncertainty that had complicated the original records. But Eleanor reminded herself that the same wording could have been copied by several people after one person introduced it.
“Does this tell us Mark wrote the sentence?” Claire asked.
“No,” Helen replied. “It tells us the sentence appeared in a version associated with the project file and that it was retained in the version prepared for circulation. The account information is not enough to assign authorship. We need a more reliable connection between the edits and the people involved.”
Helen then showed them a second change. In one early draft, the summary included a line stating that the supporting invoice had been received after the payment was released. That sentence disappeared from a later version. The final summary described the invoice as supporting the transaction without mentioning when it had been created.
Claire looked at Eleanor. “That omission makes the payment look normal.”
“It does,” Eleanor replied. “But we still need to know whether the person who removed it knew the invoice was created later.”
Helen agreed. A writer might remove a detail because it seemed repetitive, because it belonged in a separate accounting note, or because it contradicted the intended narrative. The reason could not be inferred from the edit alone. However, the combined effect of the changes was clear: the final summary omitted the unresolved queries, the absence of documented approval, and the timing issue associated with the invoice.
Martin requested that the archive provider identify any emails sent with the successive drafts. If the drafts had been circulated for review, the accompanying messages might show who asked for the changes and whether anyone questioned the revised wording. He also requested a list of the people who had access to the shared project account during the relevant period.
The archive provider replied that the shared account had been used by several staff members, but it could not immediately produce a complete access list. The project had passed through an old software system that was no longer supported. Some access records had been lost during migration. The provider could supply the names of staff assigned to the account, but it could not guarantee that the list captured every person who had used it.
Martin explained the limitation to Eleanor and Claire. The missing access history meant that they might never be able to identify the editor through the electronic record alone. Their best chance would be to compare the recovered drafts with the email chain, any handwritten instructions, and the recollections of the people who had participated in the closing process.
Claire found the limitation difficult to accept. “If the system can't tell us who changed the document, then anyone can say they didn't do it.”
“That is why we compare the records with what each person says,” Martin replied. “An uncertain electronic trail does not make the other evidence meaningless. It means we have to be careful about the conclusion we draw.”
Later that day, Robert Hales sent Martin a short supplementary statement. After reviewing the documents again, he remembered a meeting in which the team discussed whether the reconciliation certificate could be used to complete the summary before the supporting schedule had been fully checked. Robert said he had objected to describing the accounting as finally approved until the outstanding queries were resolved. He could not remember whether his objection had been recorded in writing, and he did not know who had ultimately authorized the final wording.
The statement supported the existence of a disagreement within the project team about how the certificate should be described. It did not establish that Robert’s objection had been followed or ignored deliberately. But it made the final summary’s certainty harder to explain as a simple, unanimous understanding that the accounting was complete.
Eleanor asked whether Robert remembered who had attended the meeting. Martin said he had requested that information, but Robert could only recall two other people with confidence. One was an accounts assistant who had since retired. The other was a project manager whose name appeared in the administrative correspondence. Neither had yet provided a statement.
Martin contacted the retired accounts assistant, who agreed to search a box of old papers. She did not remember the disputed payment, but she recalled keeping copies of unusual queries because she had sometimes needed to explain them after projects closed. She promised to look for any notes relating to Peter’s schedule.
For Eleanor, the possibility of another independent record was encouraging. She was also beginning to understand why the truth had remained elusive. The project’s records were spread across departments, shared accounts, personal folders, and archive boxes. The people involved remembered pieces of the process, but few could see the entire chain. That fragmentation might explain the confusion. It might also have made it easier for an unsupported claim to become accepted as fact.
Claire returned to the question of Mark’s responsibility during their drive home. She said she believed him when he claimed he had not knowingly forged anything. Yet she could not understand why he had repeated the claim that Peter approved the accounting without asking to see the actual approval. Mark had been willing to trust an administrative message when the consequences benefited a business arrangement he wanted to pursue.
Eleanor said that this was an important distinction. A person did not need to invent a false document to contribute to a misleading account. They could rely on incomplete information, fail to verify a claim, and then repeat it with more certainty than the evidence justified. That failure could be serious even if it did not amount to deliberate deception.
“But if he knew the questions were still open,” Claire said, “then it wasn't just carelessness.”
“No,” Eleanor agreed. “And we have not yet established whether he knew that when he circulated the final version.”
The next morning, Martin received a message from Susan Pike. She had found an old notebook used for tracking document circulation. She could not promise that it covered the entire period, but one entry referred to a request for a “clean copy” of the transaction summary after the outstanding queries had been discussed. The entry included a date and a set of initials, but no explanation of who had requested the copy or what “clean” meant in that context.
Susan sent Martin a photograph of the page and offered to bring the notebook to his office. Martin asked her not to remove any pages or add annotations. He arranged to examine the notebook in person and record the relevant entry before asking her questions about it.
When Eleanor and Claire heard about the notebook, neither rushed to interpret the phrase. “Clean copy” could mean a document without handwritten marks, a version with corrected formatting, or a summary stripped of qualifications. The surrounding entries would be needed to understand how Susan had used the term.
At Martin’s office, Susan opened the notebook to the marked page. The entry recorded a request for a clean copy of the summary and referred to a review of the outstanding queries. The initials appeared to match those used by a project manager elsewhere in the file, but Susan would not identify the writer without a more reliable comparison. She remembered that “clean copy” was a common phrase in the office, usually meaning a version prepared for circulation rather than an annotated working document.
Martin asked whether the phrase could also refer to removing unresolved questions. Susan said it could, depending on the instruction that accompanied it. She could not remember the particular request. The notebook did not say whether the queries had been resolved or whether the clean copy was meant to preserve the existing qualifications.
Helen reviewed the entry and said the notebook was useful because it placed a request for a new version near the period when the summary’s wording changed. But it did not prove that the request caused the changes. The initials might identify the person who requested the copy, the person who received the request, or the staff member who recorded it. The notation alone could not distinguish among those possibilities.
Eleanor felt both hope and frustration. Each recovered record brought them closer to the sequence of events, but the most important choices still seemed to occur in the spaces between documents. Someone had decided that a clean copy was needed. Someone had removed the qualification. Someone had allowed the final summary to state that approval existed without a complete record of that approval. They still needed to determine whether those were three actions by one person or separate decisions made by several people.
Martin advised them to wait for the retired accounts assistant’s search before contacting the project manager named in the notebook. The new record might provide context that would make the questions more precise. A direct confrontation without that context could produce a defensive exchange rather than a useful account.
Claire agreed, but she asked Martin to preserve the notebook photograph and Susan’s statement with the chronology. She wanted the evidence organized so that no one could later claim they had misunderstood what had been found. Martin assured her that he would keep the documents separate from the interpretations they had drawn from them.
That evening, Eleanor returned to Peter’s original letters. The handwriting was steady, the questions specific, and the requests repeated with increasing urgency. Peter had asked for the schedule, the supporting invoice, and confirmation of the approval. He had not claimed that he knew who was responsible. He had simply refused to treat the matter as settled while the records disagreed.
Eleanor understood now that this refusal was the most reliable fact about Peter’s position. Whatever else had happened, he had continued to challenge the accounting. The final summary had erased that uncertainty. Whether the erasure was deliberate or careless, it had changed the story told about him.
Before bed, Claire said she wanted to meet Mark again once the next records arrived. She did not want to make the meeting a demand for a confession. She wanted him to explain why he had been comfortable using the word “approved” when he had never seen the supporting confirmation. She also wanted to know whether he would correct the summary now that he understood its weaknesses.
Eleanor told her that those were fair questions. A willingness to correct an unsupported statement would not settle every issue, but it would reveal something about Mark’s present judgment. His response would matter even if the investigation never identified the person who first removed the qualification.
The next morning, the retired accounts assistant called Martin. She had found a copy of a message she believed related to the disputed payment. She could not yet say whether it was the original or a printed copy forwarded to her for filing. She agreed to bring it in the following afternoon.
Eleanor wrote the appointment into the chronology. The message might identify who had requested the clean copy, or it might merely repeat the instruction that the accounting should be treated as final. Either way, it was another opportunity to test the competing explanations against a record that had survived outside the main project archive.
The search had become less about discovering a single dramatic document and more about reconstructing a series of decisions. Eleanor knew that the final answer might not be simple. But she was determined that whatever they concluded would rest on what the records could support, not on the person they most feared or the person they most wanted to trust.
Click here to continue reading: PART 27: An Old Message Reveals Who Requested the Clean Copy, While Claire Faces a Choice Between Protecting Her Marriage and Correcting the Record
The Message Arrived Before Breakfast, and by Noon Her Daughter Had Changed the Locks on Every Door That Mattered
Part 26 of 35
