Several conditions, some of which are connected with the atmosphere and some with the rock, determine the efficiency of this process. Since the breaking of the rock results from the expansion and contraction due to its changes of temperature, it follows that, other things being equal, the greater the change, the greater the breaking; but the suddenness of the temperature change is even more important than its amount. It follows that great daily, rather than great annual, changes of temperature[20] favor rock-breaking, though with changes of a given frequency their effectiveness is greater the greater their range. A partial exception to this generalization should be noted. If abundant moisture is present in the pores and cracks of the rock a change of temperature from 45° to 35° (Fahr.) might be far less effective in breaking the rock than a change from 35° to 25° in the same time, for in the latter case the sudden and very considerable expansion (about one-tenth) which water undergoes on freezing is brought into play. This may be called the _wedge-work of ice_. The daily range of temperature is influenced especially by latitude, altitude, and humidity. Other things being equal, the greatest daily ranges of temperature occur in high-temperate latitudes, though to this general statement there are local exceptions, depending on other conditions. High altitudes favor great daily ranges of temperature, so far as the rock surface is concerned (see Figs. 29, 30), for though the rock becomes heated during the sunny day, the thinness and dryness of the atmosphere allow its heat to radiate rapidly at night. Here, too, the daily range of temperature is likely to bring the wedge-work of ice into play. Since the south side of a mountain (in the northern hemisphere) is heated more than the north, it is subject to the greater daily range of temperature, and the rock on this side suffers the greater disruption. Similarly, rock surfaces on which the sun shines daily are subject to greater disruption than those much shielded by clouds. Isolated peaks, because of their greater exposure, are subject to rather greater daily ranges of temperature than plateaus of the same elevation. [Illustration: +Fig.+ 28.—Exfoliation on a mountain slope. Mt. Starr-King (Cal.) from the north.] The daily range of temperature is also influenced by humidity. Because of the effect of water vapor in the atmosphere on insolation and radiation, a rock surface becomes hotter in the day and cooler at night beneath a dry atmosphere than beneath a moist one. Aridity therefore favors the disruption of rock by changing temperatures.