The agency which could produce this increased drawing northward of moisture from the trade-wind belt would be the winds blowing into the lows. According to the cyclonic hypothesis, many of these lows would be so strong that they would temporarily break down the subtropical belt of high pressure which now usually prevails between the trades and the zone of westerly winds. This belt is even now often broken by tropical cyclones. If the storms of more northerly regions temporarily destroyed the subtropical high-pressure belt, even though they still remained on its northern side, they would divert part of the trade winds. Hence the air which now is carried obliquely equatorward by those winds would be carried spirally northward into the cyclonic lows. Precipitation in the storm track on the margin of the relatively cold ice sheet would thus be much increased, for most winds from low latitudes carry abundant moisture. Such a diversion of moisture from low latitudes probably explains the deficiency of precipitation along the heat equator at times of solar activity, as shown in Fig. 3. Taken as a whole, the summer conditions, according to the cyclonic hypothesis, would be such that increased evaporation in low latitudes would cooeperate with increased storminess, cloudiness, and fog in higher latitudes to preserve and increase the accumulation of ice upon the borders of the ice sheet. The greater the storminess, the more this would be true and the more the ice sheet would be able to hold its own against melting in summer. Such a combination of precipitation and of protection from the sun is especially important if an ice sheet is to grow.