A couple of days ago I posted a diary examining the extremely thin connection the FBI established between Bruce Ivins and the Princeton N.J. mailbox where the anthrax letters were posted on two seperate dates. The F.B.I. has established another rather tenuous connection between Ivins and another key peice of evidence; the envelopes used in the mailings. And while this connection may be a little (very little, in IMHO) tighter than the Princeton nexus, it is still a stretch...
The evidence described by the F.B.I. would lead one to beleive that they have a solid connection between Ivins and the envelopes used in the mailings. And upon first reading it sounds pretty compelling:
* Envelopes used in attacks were all pre-franked sold only by post offices during 9-month period in 2001; analysis shows defects in ink on pre-printed portions of envelopes; this defect is similar to defects in printing sold by the post office in the Frederick, MD area (where Ivins lived and maintained a PO box); spokesman calls it "very likely" envelopes were purchased in Frederick MD (though does not call it conclusive or a positive match)
So how many of these envelopes were printed? About 45 million. But wait, some of these envelopes had a printing defect particular to a much smaller number of the envelopes. Just how much smaller? Either the F.B.I. could not determine just how many, or they chose not to share that information. What we do know is that these envelopes were distributed to MULTIPLE Post Offices in the Washington D.C. area from the Dulles Stamp Distribution Office in Dulles VA. Marcy at Empty Wheel has been all over this issue:
For example, Taylor made an incredibly misleading statement to suggest that the envelopes used in the attack were only available in Frederick Maryland. He claimed that, "based on the analysis, we were able to conclude that the envelopes used in the mailings were very likely sold in a post office in the Frederick, MD post office in 2001." He continued to say that Ivins maintained a PO Box "at the post office from which these pre-franked envelopes were sold."
But the truth is that Frederick Maryland is just one of hundreds of post offices at which those envelopes would have been available:
Subsequent to the attacks, an effort was made to collect all such envelopes for possible forensic examination, including the identification of defects that occur during the envelope manufacturing process. As a result of this collection, envelopes with printing defects identical to printing defects identified on the envelopes utilized in the anthrax attacks during the fall of 2001 were collected fiom the Fairfax Main post office in Fairfax, Virginia and the Cumberland and Elkton post offices in Maryland. The Fairfax Main, Cumberland, Maryland, and Elkton, Maryland post offices are supplied by the Dulles Stamp Distribution Office (SDO), located in Dulles, Virginia. The Dulles SDO distributed "federal eagle" envelopes to post offices throughout Maryland and Virginia. Given that the printing defects identified on the envelopes used in the attacks are transient, thereby being present on only a small population of the federal eagle envelopes produced, and that envelopes with identical printing defects to those identified on the envelopes used in the attacks were recovered fiom post offices serviced by the Dulles SDO, it is reasonable to conclude that the federal eagle envelopes utilized in the attacks were purchased from a post office in Maryland or Virginia. [my emphasis]
In other words, Taylor suggests, inaccurately, that Ivins' post office was the only one where those envelopes were available, rather than one of many post offices.
So, bottom line, while the same type of envelopes that were used in the anthrax mailings were available at the Frederick MD Post Office, they were also likely available to hundreds of other offices in the area. And the fact that envelopes were only found in three other offices means nothing for several reasons. First, given the time lag in the investigation, (the envelopes were sold in 2001, the investigation with Ivins as the focus in 2007) many Post Offices likely sold out of the envelopes. Secondly, having worked for the Post Office, and having been involved in Post Office window operations at various times, I can tell you that if several offices got those envelopes from the Dulles Stamp Distribution Center, then it is virtually certain that they all did.
The F.B.I. statement that it is "likely" that the envelopes used in the anthrax mailings were purchased at the Frederick MD Post Office is not just inaccurate or misleading; it is a complete and total fabrication. What we do know is that thousands, perhaps millions of these envelopes with the printing defect were distributed to many Post Offices in the D.C. area. But that does not tie Ivins to the envelopes any more than any other technician or scientist who worked at the Fort Detrick lab. All of them live in D.C. area, and all of them would have had the same access to those type of envelopes.
As with my previous diary, I have a few questions:
Of the 45 million federal eagle envelopes printed in 2001 and 2002, how many had the identified printing defect?
Just what was this defect, and how was it determined that only a small percentage of the 45 million envelopes contained said defect?
There are 89 stamp distribution centers operated by the USPS. How many of these SDCs receieved the federal eagle envelopes in question?
If other stamp distribution centers across the U.S. also received federal eagle envelopes printed by the same vendor, were any of them checked for the printing defect?
The F.B.I. states that they sampled a number of Post Offices for the federal eagle envelopes following the anthrax attacks. Exactly when was that sampling done?
There are normally procurement records kept both at individual post offices and at stamp distribution centers documenting the type and amount of stock shipped. Did you check those records, and if so, what did they reveal?