As we have already seen in “How Plants Get Their Food and Water from the Soil,” the water is the carrier of the food elements from the soil, but water as such does much more for the plant than act as a carrier. Osmotic pressure, a never-ending pump, keeps sending up a steady stream of water to the limits of its power. In everything except trees it seems fairly certain that this pressure is sufficient to drive water into the remotest leaves. It finally reaches these tiny rooms in the leaf about which we read in the account of Leaves as Factories. And just here a very curious thing happens. Each room is, as we have seen, a very busy place, crowded with all the necessary equipment to make sugar, and yet there is still room for water which is just as necessary as the other fittings; in fact so necessary is it that the whole interior of the room is bathed in water. This irrigation system works so well that the walls of the room literally bulge with the pressure of the water in them. If they did not--a condition known as turgor--the plant would at once wilt, and if no new supply came it would wither and die. But water cannot stay in this condition of pressure and stagnation for even a brief period. That would be as if a leaf were like a toy balloon which, after inflation, had the entrance pinched and so remained inflated. And while we have all along spoken of factories for making sugar, and pressure pumps for forcing up food and water, it must never be forgotten that this marvelously adjusted mechanism is a living thing. Constantly growing, even producing their own means of falling in the autumn, leaves must be thought of as living machines, just as we are still more highly developed machines. In other words the accumulated water in the cells of the leaf must be removed, after it has served its use, and replaced by fresh supplies. The removal is carried on by its evaporation into the halls, or, in the more precise terms of our account of leaves as factories, into the _intercellular spaces_. It will be recalled that these are connected with the outside air through the pores or stoma. When the air outside is hot and dry it might easily suck out by evaporation all the water vapor in these intercellular spaces and wilting follow at once. This would actually happen if the guard cells, already mentioned, were not constantly on the job. They control the size of the opening just as certainly as a steam valve does, and maintain, with a few exceptions, just the proper amount of water loss not only to maintain turgor, but to see to it that transpiration, as this process is called, goes on rapidly enough to insure fresh supplies of water being sent to the leaf. The opening and closing of the stoma by the guard cells is a nicely balanced operation dependent upon root pressure, turgor, and atmospheric conditions. Guard cells have, because of this, been much studied in spite of the fact of their microscopic size. We now