An important property rendering drying of wood peculiarly difficult is the changes which occur in the hygroscopic properties of the surface of a stick, and the rate at which it will allow moisture to pass through it. If wood is dried rapidly the surface soon reaches a condition where the transfusion is greatly hindered and sometimes appears almost to cease. The nature of this action is not well understood and it differs greatly in different species. Bald cypress (_Taxodium distichum_) is an example in which this property is particularly troublesome. The difficulty can be overcome by regulating the humidity during the drying operation. It is one of the factors entering into production of what is called "case-hardening" of wood, where the surface of the piece becomes hardened in a stretched or expanded condition, and subsequent shrinkage of the interior causes "honeycombing," "hollow-horning," or internal checking. The outer surface of the wood appears to undergo a chemical change in the nature of hydrolization or oxidization, which alters the rate of absorption and evaporation in the air. As the total amount of shrinkage varies with the rate at which the wood is dried, it follows that the outer surface of a rapidly dried board shrinks less than the interior. This sets up an internal stress, which, if the board be afterward resawed into two thinner boards by slicing it through the middle, causes the two halves to cup with their convex surfaces outward. This effect may occur even though the moisture distribution in the board has reached a uniform condition, and the board is thoroughly dry before it is resawed. It is distinct from the well-known "case-hardening" effect spoken of above, which is caused by unequal moisture conditions. The manner in which the water passes from the interior of a piece of wood to its surface has not as yet been fully determined, although it is one of the most important factors which influence drying. This must involve a transfusion of moisture through the cell walls, since, as already mentioned, except for the open vessels in the hardwoods, free resin ducts in the softwoods, and possibly the intercellular spaces, the cells of green wood are enclosed by membranes and the water must pass through the walls or the membranes of the pits. Heat appears to increase this transfusion, but experimental data are lacking. It is evident that to dry wood properly a great many factors must be taken into consideration aside from the mere evaporation of moisture. Losses Due to Improper Kiln-drying In some cases there is practically no loss in drying, but more often it ranges from 1 to 3 per cent, and 7 to 10 per cent in refractory woods such as gum. In exceptional instances the losses are as high as 33 per cent. In air-drying there is little or no control over the process; it may take place too rapidly on some days and too slowly on others, and it may be very non-uniform.