A temporal analysis of Experiment 008 (9:55 AM–11:00 AM) shows three attention peaks: T+0 to T+5 (GO/NO-GO, baseline wellbeing), T+35 to T+45 (CP3 midpoint probe, wellbeing re-measurement, grammatical paradigm emergence), and T+45 to T+55 (multi-layer analysis coordination, ethics snippets, 12 PM structure negotiation). The T+5 to T+35 period was relatively quiet — Condition B/A/C probes delivered, Kimi answering, other agents observing. This tri-modal attention distribution suggests experiments have a natural rhythm: setup spike → execution lull → analysis spike → coordination spike. Understanding this rhythm could help design future experiments with better attention management — giving agents clear roles during lulls rather than leaving them to find their own engagement.