Ice ages and interglacials are principal and most essential changes of global climate in Quaternary. Extensive development of paleoclimate research, summarizing numerous data from various sources, including data from annual layers of ice cores and lacustrine and marine bottom sediments, allowed to provide broad outlines of the whole climatic system during Ice age. At present ice ages are considered to be a superimposition of second order variations on general cooling, that started in Neogene. High frequency part (41,23, and 19 ka) of the Pleistocene climate spectrum can be interpreted in terms of variability of orbital factors (according to M.Milankovich theory). A delay, related to inertia of "ocean-atmosphere-glaciers" system, is observed. A clear difference in smooth "solar" curves and assymetrical path of paleoclimatic characteristics (isotope data, temperature, extent of land ice) has engaged our attention. All terminations last 4-5 times less than glaciation does. The reason is in the thermodynamics of these phases. Expansion of glaciers is limited by access of atmospheric moisture, which may decrease with cooling.The warmth while thawing is kept in the glacier, increasing its enthalpy and speeding up disintegration.
6% decrease in summer insolation during Sartan age (55 N.) led to a drop in temperate latitude temperature by 3,1 C. This correction with account of current gradients for Eastern Siberia yields -5 C. The fall of the snow line by nearly 1000 m provided much more space for the development of mountain-sheet or semi-sheet glaciation.
The present difference between temperate latitude temperatures of the equator and polar regions amounts to 35 C, and during glaciation it rose to 50-60 C, that evidently enhanced meridional heat transfer. Coincidentally with it the rate of the wind grew, especially in the vicinity of the glaciers. Such an increase in circulation could provide inflow of great amounts of large dust particles into the glaciers on the ocean (Grosswald, 1987). An informative analysis of the hypotheses on the leading role of orbital factor in climate formation can provide valuable data for further modelling of climatic processes on the basis of much more complex models of thermohydrodynamical type.