The most damaging evidence I've seen thus far linking Dr. Bruce Ivins to the anthrax mailings is his uncharacteristic spike in night work in the days leading up to both mailings. I've also seen a lot of mischaracterization of this time in both directions.
Let me make my position clear from the beginning. We don't have all the evidence we need to conclude that Ivins is the anthrax mailer, nor that he was part of a group that did this. The FBI released a great deal of information in its release of search warrant affidavits, and I think they had a case for probable cause there. But contrary to their assertions, I've not seen anything like enough evidence to overcome a reasonable doubt defense, not if I was in the jury box. And the work hours, while the strongest evidence the FBI has got, is still just not enough.
In fact, their own release of evidence shows that there's more to the story that they haven't revealed.
Let's start out with laying out exactly what the FBI has released concerning Ivins' spike in night work in August, September, and October of 2001. First the chart demonstrating the spike:
This lays out his work hours at night in the B3 lab the years 2001 and 2002. As anyone can see, there is a definite spike in August, September, and October of 2001.
Furthermore, an approximate number of hours can be taken from this chart. In August, Ivins worked about 11 hours and 30 minutes (11:30) in B3 at night (after 6:00 p.m.). In September, he worked about 32:30, and in October, 16:30. These three months combined could almost rival the 19 previous months combined in hours.
Remember these times.
Next, the FBI released the times for Ivins' late night shifts on the three days before the first anthrax mailing:
Not only are the times Ivins was in the B3 lab released (the fourth column), but his actual times in the building housing the lab are also released, down to the minute. The two shouldn't be confused.
So on September 14, 15, and 16, Ivins worked 2:15 each day in the B3 lab, a total of 6:45. The anthrax mailing according to the FBI happened in a window from 5:00 p.m., September 17 to noon, September 18.
Finally, the last piece we'll look at here, the times released by the FBI of Ivins' night work in B3 before the second mailing:
More more time here (fitting with the much more refined nature of the anthrax mailed the second time). Here, Ivins' hours in B3 are totaled as 4:20 in September and 10:57 in October. These were incurred in eight consecutive nights from September 28 to October 5. The window of the second anthrax mailing was from 3:00 p.m., October 6 to noon, October 9.
The total time Ivins spent in B3 in October was 16:30, so in the first five days of October, he racked up the majority of that time, 10:57. That leaves 5:33 for the entire rest of the month. That's not so good for Ivins.
But the September numbers are different. Between the two tables, Ivins' time in September that we know of totals 11:05 (6:45 for the first table, 4:20 for the second). However, this was the big month, where Ivins racked up 32:30 in the B3 lab at night. So both of these tables leave 21 hours and 25 minutes unaccounted for in the month of September.
When did these hours happen? We don't know, but we can certainly narrow it down.
Between the first table and the second, there's a gap from September 18 to 27. The FBI's search warrant gives us an idea of this time period:
He wasn't in there. Again this doesn't look good for Ivins. But knowing that the second table is consecutive days, we can also rule out September 27 (or the FBI would have included it, making their probable cause case stronger). So on September 25, Ivins was back in B3 at night, and he might have been on September 26 as well.
But these dates couldn't have been very substantial time periods or I think the FBI would have thrown them into the mix as well. It's likely to me that Ivins was only in the lab 15 minutes or so on the 25th, and not at all on the 26th, following the pattern after the second table (and thus making the case as strong as possible for the FBI).
No matter how you figure those two days, you cannot escape the fact that the bulk of Ivins' September time in B3 at night happened before the first date on the first table, September 14. That's thirteen days with around 21 hours of B3 night hours logged.
And there is no accounting for the August hour spike at all by the FBI. We simply don't know how Ivins spent his time in the B3 lab that month.
Very likely, if the FBI had released a complete table of Ivins' August, September, and October 2001 night hours in B3, the very damning pattern they established in the search warrants would have been muddied. If it had helped their case, it would have been part of these papers.
Ivins' lawyer has given this explanation for that time:
...we never knew what the exact records were, but he'd been questioned about this, was questioned about it in the grand jury, gave them his best recollection, that he was having various family problems, both in his marriage, and concerning one of his children, at the time. And because of that, was having trouble getting work done during the day, went over there, was doing work at night, as he always did. And they had a chart showing he was always going in there, he just went more frequently during September and October, and then continued to do so right through the end of the year.
The chart at the top shows that Ivins was indeed in the habit of going into the B3 lab after 6 p.m. off and on. And notice what the lawyer said: "continued to do so right through the end of the year."
The FBI's chart doesn't bear this out, however. The November and December 2001 hours do trend down from the spike in September, but they are both in the range of hours found in the 19 months before August 2001.
Indeed, from the last sentences quoted from the FBI search warrant (dated Oct 31, 2007), we have 1:41 more accounted for in October, making the total we don't know about 3 hours and 52 minutes for October 15-31. That doesn't support what the lawyer is saying, and fits in with Ivins' normal pattern (before August 2001).
Still I would be surprised to learn that in the days after the second anthrax mailing, someone could easily and for long periods of time enter the B3 lab at night like that. Perhaps so. But I would think that things would have to have settled down before the unit could get back to normal. And the question of the spike in time for August and the first two weeks of September remains unanswered.
The extent of Ivins' troubles at home should also be nailed down. If this actually did start in August 2001 and continue on past the anthrax mailings, I would expect other spikes to begin to manifest in the days after the unit got back to a semblance of normalcy. If, however, the "normal" pattern continues, the late 2001 spike would be all the more damning.
This evidence remains very bad for Dr. Ivins. It is not, however, the total slam dunk that the FBI is presenting it to be. The FBI needs to be more forthcoming and release the total amount of data they have, including information on Ivins' work hours from August through at least 2003.