Many of the old powders start a fast pressure spike when they reach a certain pressure. In other words, as powder is added to increase velocity, and pressures climb with each increase in charge, gradually come to that pressure spike threshold, then pressures rise much more rapidly with each increase in powder added. H 110, 296, and Accurate 9 are all ball powders which have a straight pressure climb all the way to 100,000 psi. 2400, and I surmise from my experiance, Unique, are two that go on a pressure excursion when they hit a certain ceiling. As I recall, the ceiling for 2400 is something around 50,000. If anyone is sure of what it's ceiling is, please set the record straight. That number is only something that comes from a foggy memory.
Having the exact numbers is not the concern here, but only that reloaders realize such a thing exists, and watch for the effects when building up heavy loads. It shows on a chronograph quite dramatically. As a hypothetical example, suppose each .1 grain increase of a powder was increasing velocity by 30 fps, when at a certain point velocities began to go up by 40 or 50 fps per .1 gr of powder increase. That's the point where one should try working with another powder.
A load that is at this threshold, and developed in cool weather, will go way over the threshold if shot at considerably higher temperatures. For example I once loaded up my deer chasing 243 ammo at daybreak, when temps were cool, for Arizona, then when I got out hunting and temps climbed to near 100, primers were severely flattened and cratered, and the bolt began to open with difficulty. That was when I learned of the pressure excursion problem, back about 1974 as I recall.