The stories are all over the place.
ABC says some intereseting things,
http://abcnews.go.com/TheLaw/FedCrimes/Story?id=5603993&page=2
...The scientists noted that the anthrax used in the attacks had no additives on the anthrax spores, but that the mineral silica was present in the deadly substance.
Although the FBI was able to reverse-engineer anthrax similar to the anthrax used in the mailings, scientists have been unable to reproduce it with the silica. ...
...As the FBI constructed a genetic fingerprint of the anthrax used in the attacks, investigators found that the second set of samples Ivins provided to the bureau had none of those genetic markers, but that the sample called RMR-1029 matched the four key genetic markers.
http://www.msnbc.msn.com/id/26279202/
Msnbc
..FBI officials admitted Monday that destroying the initial Ivins sample was a mistake, but said it didn’t really hinder the investigation because the technique used to trace the source of the anthrax to Ivins had not been developed yet. Luckily, a copy of that first sample was sent to an outside professor, who years later used it to help further link Ivins to the killer strain...
(But it sounded better as innuendo that Ivins’ acts delayed the investigation)
...Contradictions in accounts
At times the officials and scientists contradicted themselves, even down to the number of flasks containing the anthrax Ivins had. They eventually agreed it was one one-liter triangular flask capped with cheesecloth that linked Ivins to the attacks...”
...FBI officials spent much of their time going into more detail about the specific four mutations in the strain of anthrax that led back to Ivins’ Army lab at Fort Detrick, Maryland. They were not easy to find, rare like red or green M&Ms in a flask mostly full of blue candies, Majidi said....
(terrible metaphor)
In February 2002, the FBI sent subpoenas around the world to labs handling the Ames strain of anthrax, which was the strain that killed five people in 2001. They got 1,070 samples and destroyed only one: the first one from Ivins. It was destroyed because Ivins didn’t use the proper test tube and growth medium so it may have not been useful as evidence in court, officials said.
Ivins was one of the first to respond to FBI subpoenas, but his sample was deemed useless and he was asked to submit another. He gave investigators a second sample of anthrax from his lab in April 2002 to comply with standards in a subpoena issued in the case. But that sample contained a different strain than what he submitted two months earlier in what prosecutors call an attempt to deceive or confuse investigators...
(hmm...maybe the request said “Ames” and Ivins had more than one sub-strain, so the fact IF he sent two different “samples” neither act violated a request for “Ames strain”.)