I've never come across any brass that would string like that. I havn't used enough to say it "can't", but I've never seen it. Bronze on the other hand..... I've found some stringy bronze chunks in my rummaging.... Unfortunately, what worked best for me on "regular" cutters was no top rake (flat top), and a mid to light amount of of side rake. No bueno for a parting tool, they're already high strung enough as it is, they don't need more excuses to complain.
So I SUSPECT that you're into a bronze here, not a brass. The tool you're using does appear to be "cutting" that chip, not "buldozing" it. The second batch of chips- I see inconsistancy there, like maybe hesitant infeed (or just tired out hamds after half an hour....), but I'd call that general result a good direction to follow.
After that, because I think the tool is actually cutting, (not buldozing), that would make the pertinant variables being the surface speed, and the infeed rate. What were you doing when the descrete chips formed?
If you could speculate the RPMs and the actual diameter of the cut at any particular point where it was going better (not the whole part, the point where the tool was touching the work, that's where surface speed is), that might give you a good idea as to the surface speed that your lathe and your tool was working closest to "working well".
And the infeed- That changes with diameter too, but not as much. At least by feel, if you're hand feeding.. That wants to be smooth, steady, and constant. (Because "constant" is more about handwheel effort and less about actual infeed per revolution...) Load the tool up and keep it that way. If you need to stop, don't just stop, reverse and pull the tool back a whisker.
I think (I don't know, I think) that those two items there (and basically some practice and experimentation) is going to solve the problem here. The speed (which is surface speed of the work, NOT the rpm of the machine) will change dramatically as the diameter reduces. The infeed will reduce slightly as the diameter reduces, especially towards the center. (well, they both blow up into a singularity at the center, but by "feel", that's how they act).
So, when you have the opportunity and material to practice that again, keep your notes. Work out the surface speed of the work, infeed at one rate that you set with your hands, (one pressure on the hand wheel). Add some speed. Half the speed. Keep the infeed the same. keep all those notes. If faster or slower is good, put it in the noties. Try more yet. Or split the difference. Split it again. Find a good speed for you and yoru equipment and the material at hand. Then play with the feed. Same thing. A whisker more force. A whisker less force. Creap up on it. Give it the beans. Split the difference. Keep it steady...
It's a steep learning curve going in, but it starts to level off pretty quickly once you get going. Why so steep at first? Because you have to come to realize that your parting tool hates you. Never forget that. You will ALWAYS have to stay on top of it, to keep it acting properly. But with familiarity it gets a LOT easier.
So why don't I give you numbers? Well, because numbers are dependant on YOUR machine and YOUR tools. And after the steep part of the learning curve, you're not going to use too much for numbers either. Surface speed to start maybe. But you're going to find that the parting tool, with two constantly changing variables that change in different ways, is driven more by feel than it is by hard numbers. What I'm suggesting that you document EVERYTHING at every cut. That's to not to start a database, it's to establish a mental picture in your head. It's not forever, and looking up notes is NOT the final solution. You need -your- numbers to get started, it'll seem like a nightmare at first, but like I say, the steep inital learning curve quickly levels off, and muscle memory takes over, and the math reduces to just figuring a starting speed. Feed per inch reduces to an accurately appplied amount of hand pressure, and the ever reducing surface speed will reduce from calculating what diameter needs a gear change at what diameter to "yeah, that's slow enough, I can give it another gear now"
I said no numbers. But just for jumping off the cliff..... You're "probalby" cutting best (with "regular" cutters) at half to two thirds of the official numbers for "production work". So take whatever speed YOU are efficiently cutting the material at, and the parting speed will "probably" be half to two thirds of that number.
Zero guarantees, and don't be afraid to step outside of those estimates, high OR low. Just do it methodically, and not randomly.