The earth alternates in a supercycle 400 m.y. long from a single continent (Pangea) and ocean (Panthalassa) with an icehouse climate to many continents and oceans with a greenhouse climate. The supercycle is driven by the heat that accumulates beneath the insulator of Pangea, and the greenhouse is made by the excess CO2 vented from the mantle during the faster rate of plate activity. In the current supercycle (A) since the mid-Carboniferous (320 Ma), the growth and decay of Pangean heat is observed in globally synchronous cratonic sedimentary sequences that represent, in ascending order: (1) a stratigraphic gap, (2) sagging, (3) rifting, (4) fast, and (5) slow seafloor spreading. Stages 4 and 5 of the previous supercycle (B) are observed back to the start of the Phanerozoic; in the Proterozoic, earlier stages of B are postulated back to 720 Ma and a third supercycle (C) to 1100 Ma. The big glaciations of the latest Proterozoic (700-600 Ma) and Late Palaeozoic (320-260 Ma) are confined to the icehouse state during stages 1-3; the Sturtian (800 Ma), latest Ordovician (440 Ma), and Quaternary (2-0 Ma) glaciations fall within greenhouse states during stage 5 of slow seafloor spreading. The next icehouse glaciation is expected 80 Ma hence. The problem of long-term past (and future) climate change is soluble by global studies. Urgently needed are (1) global stock-takes of (a) the rate of granite emplacement, which serves as a proxy via plate activity for the mantle source of atmospheric CO2, and (b) carbonate sinks and other sedimentary indicators of CO2; and (2) the extension of the sea-level curve back past 570 Ma to the start of the plate-tectonic eon 1100 Ma ago. From such studies, environmental baselines for past climatic states can be drawn for gauging present and future "background" variations.