Streak plate and Gram stains don;t give confidence of pure culture(s).

Consult Bergey's Manual for Determinative Bacteriology. Generally, anything that forms chains of cells, which is how I am interpreting your description of morphing into chains, is a yeast, mold, or fungi, not a bacteria, but that is not a hard, fast, rule. Gram stain is a first cut Gram + from Gram -, but within each category are tens of thousands of possible species. Cellular morphology is the next discriminator. To get down to the Genus and species level, however, requires biochemical testing (i.e., what can the organism grow on and what chemicals can it tolerate or convert).

I had to design a Microbiological testing program for a GMP clean room years ago and found that most microbiological approaches were done serially, which just doesn't work in a production environment (i.e., takes weeks and then requires product recalls), so I devised a parallel replica plating strategy that worked in 5 days. Costly, sure, but no where near as much as a product recall. Plus, we could reduce the problem to a simple algorithm.

Looks like Bacillus to me. It even forms endospores.