I pulled the papers back from the trash. The notation was my summary of the sale of one of the real estate parcels at Paradis sur Terre. I’d listed the sale price as $2.4 million, but I felt certain the settlement sheet I’d seen in The Lab listed it as $4.1 million. Maybe I was remembering wrong, but I didn’t think so. Or maybe I’d made an error in the entry, either here or downstairs. But again, I didn’t think so.

I wasn’t allowed to take any documents out of The Lab, but nobody said I couldn’t take any documents in. I photocopied my handwritten index, and when Lester called to let me know he’d arrived, I tucked the pages inside the waistband of my skirt. They crackled a bit as I walked downstairs to The Lab, but Lester didn’t seem to notice. He swiped his keycard through the lock and let me in.

I set him to work Bates-stamping a few thousand pages that I’d already cleared for production to Mark Ivins. Once he was occupied with the machine in the corner of the room, I pulled the photocopies from my skirt and smoothed them out on the table. Then I sat down at the computer and opened the document index I’d been creating here in The Lab. It took only a moment to confirm that my recollection was correct. My handwritten notes showed proceeds of $2.4 million for that particular parcel of real estate. The entry on the computer screen showed proceeds of $4.1 million.

I wheeled my chair across the room and located the settlement sheet for that transaction. The sale price that appeared there was also $4.1 million. Okay, I thought, so I made a mistake in my handwritten notes. But I didn’t believe it.

I did a side-by-side comparison of the index on the screen against my handwritten index. I went through every asset sale from PsT—real estate, equipment, raw materials, and finished goods. In nearly every instance, my notes from the original documents in the file boxes showed a much lower sale price than the copies in The Lab. A forklift sold for either $15,000 (per my notes) or $25,000 (per the computer screen). A sewing machine for either $1,495 or $2,799. It went on and on. Cafeteria tables, trucks, air compressors, mattresses, cutting machines—all the sales proceeds were greater than what I’d recorded from the original documents in the file boxes.

I scooted my chair across the room to the table where the closing binders and bills of sale were stacked, and I double-checked the sales figures against those documents. They all matched the numbers I’d entered into the computer; so my mistake, if I’d made one, hadn’t been made in The Lab.

But it wasn’t one mistake. It would have been a mistake in recording the sale prices of literally hundreds of assets.

One mistake I could admit, but two hundred? No.

The documents in The Lab were copies. The documents I’d indexed from the file boxes were originals. The conclusion was inescapable. Someone had altered the copies to inflate the sales proceeds of the PsT assets.