30 Slirhikage, SivelUng, Warjnng of Cross-grained Woods 



by gain or loss of the same amount of water takes place when the wood 

 is poor in water, and that as the wood contains larger amounts of water 

 there is a decline in such a change in area. This is in part due to the 

 larger proportion of water present in the lumina of the wetter wood; 

 moreover if we assume that the same loss of water always causes the 

 same decrease in volume of the solid wood-substance (cell-wall) the 

 smaller diameters of the wood-constituents in the dry phase have to 

 be considered. But the irregularities in changes of area near the 

 beginning and conclusion of the drying and soaking require explanation. 

 One source of misinterpretation is brought out by the change between 

 July 16 and 19, when there was a slight hygroscopic absorption of 

 water and yet a decrease in area : probably this additional moisture 

 was superficially distributed to a considerable extent, and the main 

 mass of the wood was probably drier on the later date. During drying, 

 small changes in weight of the board may on the one hand be associated 

 with relatively greater changes in the distribution of moisture within 

 the board ; moreover during such small changes a slight underestimation 

 of the amount of water lost will cause the increase in area to appear 

 excessive. And it will be noted that during drying apparently unduly 

 large results are associated with small changes of weight, viz. on July 7, 

 11 to 12 a.m., July 12-13, July 13-15, and July 15-16. On these dates 

 the maximum result (dividend) appears when the loss of water recorded 

 was the minimum (July 13-15). During soaking there w^ere no such 

 small changes in weight, and, excepting at the beginning and end, no 

 sudden or great changes in the rate of dimensional increase. The 

 contrast between the results between June 21 and 22 at 10 a.m., and 

 June 22, 10 a.m. to 2.15 p.m., is possibly due to the fact that the increased 

 water-content during the former period was due to slow absorption of 

 aqueous vapouT, and during the latter period to rapid intake of liquid 

 water much of which would be in the lumina and in operation on 

 swelling; but the unexplained smallness of the increase in width of 

 the middle of the board on June 22 is also partly responsible. 



Twisting of Yang 1. 



The warping along the length and transversely across the ends 

 have already been discussed; it remains to consider the warping along 

 the diagonals and the consequent twisting. 



The deviations from the straight were measured in millimetres on 

 the curves recorded by Professor Dalby's instrument; but as the 



