In March 2002, Rusi Taleyarkhan, a scientist at Oak Ridge National Laboratory, announced he had produced nuclear fusion using nothing but sound waves, a claim that would upend decades of multibillion-pound fusion research if true. His method relied on sonoluminescence, a process that collapses a bubble so violently it flashes light and, he argued, briefly reaches temperatures hot enough to fuse atoms, a phenomenon nicknamed the star in a jar. Colleagues Mike Saltmarsh and Dan Shapiro reran his experiment and found nothing, while Taleyarkhan's team published a second paper claiming even stronger results. Horizon commissions its own test, asking Seth Putterman, a leading sonoluminescence researcher, to replicate Taleyarkhan's exact setup and measure neutrons against light flashes with nanosecond precision, timing fine enough to prove or kill the fusion claim outright. The film follows the panel of UK experts reviewing Putterman's data and the negative result that follows, alongside Taleyarkhan's response that equipment differences explain the discrepancy. The dispute stays unresolved, a case study in how hard it is to verify extraordinary claims in a lab.