Last week, the FBI selectively leaked the news that Bruce Ivins had taken personal leave on September 17, 2001; they seemed to be arguing that Ivins had taken leave to drive to Princeton, all in time to return for an appointment that evening.

A partial log of Ivins’s work hours shows that he worked late in the lab on the evening of Sunday, Sept. 16, signing out at 9:52 p.m. after two hours and 15 minutes. The next morning, the sources said, he showed up as usual but stayed only briefly before taking leave hours. Authorities assume that he drove to Princeton immediately after that, dropping the letters in a mailbox on a well-traveled street across from the university campus.

But then some DFH bloggers pointed out that that theory was impossible.

It would not be possible for Ivins to have mailed the anthrax. According to my calculations above, the window during which Ivins could have put the letter in the mailbox on September 17 was from 10:25 to 1:35. But here’s what the FBI itself says about the window in which the letter was mailed:

The investigation examined Dr. Ivins’s laboratory activity immediately before and after the window of opportunity for the mailing of the Post and Brokaw letters to New York which began at 5:00 p.m. Monday, September 17,2001 and ended at noon on Tuesday, September 18, 2001. [my emphasis]

In other words, had he mailed the anthrax when they’re arguing he did, the letter would have been picked up at the 5:00 PM pick-up (if not an earlier one–often boxes have a mid-day pick-up as well), and post-marked on September 17, not on September 18.

So, after their operative theory couldn’t even withstand the scrutiny of the DFHs in the world, they revised their theory.

Meanwhile, government sources offered more detail about Ivins’s movements on a critical day in the case: when letters were dropped into the postal box on Princeton’s Nassau Street, across the street from the university campus.

Investigators now believe that Ivins waited until evening to make the drive to Princeton on Sept. 17, 2001. He showed up at work that day and stayed briefly, then took several hours of administrative leave from the lab, according to partial work logs. Based on information from receipts and interviews, authorities say Ivins filled up his car’s gas tank, attended a meeting outside of the office in the late afternoon, and returned to the lab for a few minutes that evening before moving off the radar screen and presumably driving overnight to Princeton. The letters were postmarked Sept. 18.

Notice this time they’re not going to tell us what time Ivins went back to the lab, perhaps because they want their theory to hold up through the weekend. If it was late at night, it would introduce the same problems of timing, as presumably Ivins was back at the office at 7:30 AM the next morning, as per usual. And there’s one more critical detail: Ivins did not enter Suite B3 when he returned to the office that night. As the original search warrant attachment laid out,

The investigation examined Dr. Ivins’s laboratory activity immediately before and after the window of opportunity for the mailing of the Post and Brokaw letters to New York which began at 5:00 p.m. Monday, September 17,200 1 and ended at noon on Tuesday, September 18, 2001. Beginning on Friday, September 14, Dr. Ivins worked the following three consecutive evening shifts prior to the mailings with time spent in Suite B3:

Friday, September 14, 8:54 p.m. to 12:22 a.m., 2 hours 15 minutes

Saturday, September 15, 8:05 p.m. to 11:59 p.m., 2 hours 15 minutes

Sunday, September 16, 6:38 p.m. to 9:52 p.m., 2 hours 15 minutes

After September 16, Dr. Ivins did not enter Suite B3 in the evening again until September 25. [my emphasis]

As I understand it, the attachment considers anything after 4:45 to be evening work, and we know that Ivins was in an appointment at that point, so his brief return to the office must have been an evening visit. This does not, by itself, doom their new theory. After all, Ivins was caught that December cleaning his office space, not B3, so it’s possible he just stored the anthrax outside of B3 after he cultured it, if in fact he did culture it. But that would raise the question of why the anthrax didn’t show up in sampling of his office, and how and where it was stored during the day of September 17 when Ivins was not in his office.

The WaPo also notes that the hair samples in the mailbox don’t match Ivins’ hair.

Federal investigators probing the deadly 2001 anthrax attacks recovered samples of human hair from a mailbox in Princeton, N.J., but the strands did not match the lead suspect in the case, according to sources briefed on the probe.

I actually don’t think that’s a big deal–there’s no reason to believe that the culprit necessarily left hair samples on the envelopes and that no one else did, particularly since it took ten months to actually find and sample the mailbox in question.

I’m most amused by the way they find evidence, claim it’s significant, then discard it entirely once their theory falls apart. Just last week, the fact that Ivins had taken leave for the entire day before the anthrax was mailed was another smoking gun, proof that he had done the deed. But now that we DFHs have proved he couldn’t have been mailing the anthrax at that time, then that piece of evidence is discarded entirely. It seems they don’t want to entertain the possibility he was doing something else–perhaps meeting with someone, but not driving to Princeton. Similarly, lie detector tests were critical, until they discovered their main culprit passed his.

In other words, they’re still making it up as they go along, and they still don’t have a solid case that Ivins was the sole culprit.