In the dense swarm of old stars known as the globular cluster M15, astronomers have spotted not just stars but layers of time. Some stars were born earlier, others later, and their chemical composition reveals their origin with the precision of an archaeologist reading the strata of an ancient settlement. This is a story about how an explosion that enriched the gas with heavy elements left uneven traces in the first generation of stars, while by the second generation everything had mixed.

Traces of an ancient explosion, frozen in two layers of stellar population, are what the authors studied. They took spectra of 89 stars in M15 obtained with the Keck II telescope and measured the abundances of several elements produced in the r-process, the rapid capture of neutrons by nuclei. Magnesium, yttrium, barium, lanthanum, and europium: these names sound like a list of clues. It turned out that among first-generation stars, which are rich in magnesium, the spread in barium, lanthanum, and europium abundances is significantly larger than among second-generation stars. For example, the standard deviation of [Ba/Fe] in the first generation is 0.26 dex, while in the second it is only 0.13 dex. It is as if the first layer of paint was applied unevenly, while the second spread out smoothly.

What does this mean for our understanding of globular clusters? They were once considered simple systems: all stars born simultaneously from a single cloud. But observations over the past decades have shown that almost every cluster contains at least two groups of stars differing in light elements. Now it turns out that heavy elements are also unevenly distributed, and specifically in the first generation. This supports a scenario with multiple episodes of star formation: enough time passed between the birth of the first and second generations for the gas to mix. The authors used a detailed spectroscopic analysis that accounts for the effects of carbon and oxygen, making the conclusion about the difference in dispersions more robust than in studies based on photometry or smaller samples.

The outlook is broad. If this pattern is confirmed in other metal-poor clusters such as M92 and NGC 2298, it will become a strong argument for the multiple-population model. The question remains: which specific r-process source, a magnetorotational supernova, a collapsar, or something else, has the right delay to enrich only part of the first-generation gas? Answering it will refine our understanding of the early Universe and the birth of stellar systems.