Grasses transpire less freely than trees, but their root system is much shallower and less well-developed. They depend upon the water in the upper layers of soil, and must have frequent, even if gentle, showers during their growing season, while they are quite indifferent to drought and even to cutting winds during their resting period. A little reflection will show that it results from these facts that woodland, grassland and desert do not form a continuous series. It may quite well be that woodland passes through scrub into desert without the intervention of grassland. Right across Europe there is (or was) a broad belt of forest. Southward towards the Mediterranean this thins out into a characteristic form of scrub, called maquis, whose characters we shall describe later, and this scrub passes in all directions into desert land. Here no belt of grassland intervenes, for the rainless Mediterranean summer makes the growth of grass virtually impossible, except where special conditions, _e. g._ hills, introduce modifications. Contrasted with this we have the conditions in North America where, _e. g._ in Canada, the western coast is densely forest-clad, as is also the eastern region. In journeying eastward after crossing the Rocky Mountains the forest dies away into grassland, and the same thing happens, though more slowly, in a westward journey. The reason is that in this case there is a steady diminution of precipitation on passing to the interior, but what precipitation remains is, as we have seen, largely, though not wholly, summer rain, and is, therefore, sufficient to determine the growth of grass, though not of trees. Again, in North Africa the forests of the Atlas Mountains pass directly, without intervening grassland, into the Sahara desert, but to the south of the desert the grassy and park-like Sudan separates the desert from the luxuriant tropical forest. In the latter case, however, it is possible that man's influence has counted for something. On mountains, in whatever latitude, the conditions are much more uniform, partly because it is wind, assisted by temperature variations, which is the dominating factor. Moisture is usually abundant, but high up what is called physiological drought occurs; that is, the temperature is too low for the plants to be able to absorb the abundant water.