It was possible, Berger considered, he used the roof then, too. She wouldn’t rule it out, and she might never know.

Morales was smart as hell. He’d finished Dartmouth, Johns Hopkins. He was a sadistic sexual psychopath, perhaps the most outrageous and dangerous one Berger had ever come across. She thought of the times she’d been alone with him. In his car. In Tavern on the Green. And in the Ramble, when she’d paid a retrospective visit to that crime scene where the marathon runner had been raped and manually strangled, and now Berger had to wonder about her. Did Morales kill that woman, too?

She suspected it. Couldn’t prove it. A jury wasn’t likely to trust any identification based on the sound of his voice, which, like O.J. and the bloody glove, could be altered on demand so he didn’t sound exactly like the murderer in the recordings. That man spoke with a heavy Spanish accent. Morales, when speaking normally, had no discernible accent. A case wasn’t going to be won solely based on forensic voice analysis, either. Didn’t matter how sophisticated the software.

It wasn’t likely anyone—certainly not a prosecutor as seasoned as Berger—was going to suggest anything as ridiculous as making a comparison of Morales’s penis with the penis in the video recordings, a normal penis, uncircumcised, nothing unusual about it, nothing remarkable one way or another, and wearing a condom over it was reminiscent of someone having a stocking over his face. Were there any identifying features, so much as a freckle, they were masked .

The most the cops could do—or Lucy could do—was prove these violent, seemingly damning videos were in his e-mail account, but where did he get them? Having them didn’t prove he’d killed anyone or even done the filming with a camcorder he must have set up on a tripod. Lucy was the first to say that getting jurors to understand IP addresses, machine access codes, anonymizers, cookies, packet sniffing, and about a hundred other terms that were part of her easygoing vernacular was like a throw-back to the early days, the late eighties and early nineties, when people like Berger were first trying to explain DNA to judges and jurors.