Drying oj Tmiher. 69 



too rapidly, then the wood is very apt to warp, especially if the drying 

 be at all uneven. This appears to be the case with elm (Ulmus cam- 

 pestris). On the other hand, if the moisture be lost very slowly, any 

 attempt to hurry up the drying will lead either to warping or cracking, 

 as in the case of oak, where the medullary rays tend to open out. 

 Wood that contains a low percentage of moisture is not difficult to 

 season. 



We have already seen that different species contain different 

 amounts of moisture. Hence we cannot compare their rates of drying. 

 However, moisture is lost by diffusion, and we know from the laws 

 of diffusion that the amount of moisture lost is proportional to the 

 gradient of the concentration, and the area of the evaporating surface. 

 It is expressed mathematically as — 



clx/dt=Dd C.A/dl. 



Where for timber x is amount of moisture lost In grams, t is time in 

 days, c is concentration of the moisture, Z is length in inches, A is 

 area of diffusing surface in sq. inches, and D is the diffusion constant. 



The amount of moisture diffused will also be affected by Tempera- 

 ture and Humidity, but these may be omitted if the conditions are 

 kept constant. 



If this formula can be applied to the drying of wood, then the 

 value of D will give us a measure of the ease or difficulty with which 

 any particular timber parts with its moisture. Owing to the many 

 difficulties in applying such a formula to timber drying, it was not 

 expected to obtain any very accurate results. In fact, accuracy is 

 impossible. But what might be expected is that values would be 

 obtained which would give an indication of the relative powers of 

 diffusion of the various timbers. For the experimental work, blocks 

 of straight grained, freshly felled timber were cut into sizes approxi- 

 mately 2x2x3 inches. The 2x2 face was tangential and the length, 3 

 inches, was in a radial direction. The four 2'' x 3" sides were coated 

 with the mixture recommiended by Tiemann(2). The blocks were placed 

 in an oven at 40°C. and at a 50 per cent, humidity. They were weighed 

 at the same time every day, and the loss of weight plotted so that 

 any irregularity of loss might be noticed. While the blocks were 

 still actively drying, the sides were cut off and the blocks split up into 

 one quarter of an inch sections. These sections were at once weighed 

 and then dried, and the moisture percentage calculated to the dry 

 weight. These percentages were plotted as in fig. 4, and the moisture 

 gradient was obtained by drawing a tangent at the extremity of the 

 curve. The value of the gradient was substituted in the equation for 

 dC/dl. Half of the amount lost in the previous 24 hours was taken 

 as the value of dx/dt. 



The value of D obtained from the formula is by no means 

 accurate, as the values obtained are somewhat wide. This was 

 expected for various reasons. The width of ring varies greatly, even 

 in the same specimen, and may even vary widely in two adjacent 

 rings. In a pile of Red Oak (Quercus rubra) which was ready to go into 

 a kiln the width of ring varied from one-half to one-tenth of an inch. 



