It is, however, the least satisfactory of the two where careful drying is required, since the conditions cannot be so well regulated and the temperatures and humidities are apt to change with any change of wind. The apartment method can be arranged so that it will not require any more kiln space or any more handling of lumber than the progressive type. It does, however, require more intelligent operation, since the conditions in the drying chamber must be changed as the drying progresses. With the progressive type the conditions, once properly established, remain the same. To obtain draft or circulation three methods are in use--by forced draft or a blower usually placed outside the kiln, by ventilation, and by internal circulation and condensation. A great many patents have been taken out on different methods of ventilation, but in actual operation few kilns work exactly as intended. Frequently the air moves in the reverse direction for which the ventilators were planned. Sometimes a condenser is used in connection with the blower and the air is recirculated. It is also--and more satisfactorily--used with the gentle internal-gravity currents of air. Many patents have been taken out for heating systems. The differences among these, however, have more to do the mechanical construction than with the process of drying. In general, the heating is either direct or indirect. In the former steam coils are placed in the chamber with the lumber, and in the latter the air is heated by either steam coils or a furnace before it is introduced into the drying chamber. Moisture is sometimes supplied by means of free steam jets in the kiln or in the entering air; but more often the moisture evaporated from the lumber is relied upon to maintain the humidity necessary. A substance becomes dry by the evaporation of its inherent moisture into the surrounding space. If this space be confined it soon becomes saturated and the process stops. Hence, constant change is necessary in order that the moisture given off may be continually carried away. In practice, air movement, is therefore absolutely essential to the process of drying. Heat is merely a useful accessory which serves to decrease the time of drying by increasing both the rate of evaporation and the absorbing power of the surrounding space. It makes no difference whether this space is a vacuum or filled with air; under either condition it will take up a stated weight of vapor. From this it appears that the vapor molecules find sufficient space between the molecules of air. But the converse is not true, for somewhat less air will be contained in a given space saturated with vapor than in one devoid of moisture. In other words the air does not seem to find sufficient space between the molecules of vapor.