45 



dimension of their elements seem not to improve the eoiiduetivity to any grent extent. This is 

 quite (iiflfereut in a taiiRcntial tlireetion where every ek-meiit possesses dolli'd pits. But llirsc ])ils 



are .yiuuded by iricmhranes, and any one who lias watched a thin section licsidy cut IV a drv 



stick with its man.y jiersistent air bubbles, appreciates why water (h)es not [lass readily throu;;li 

 wood across the grain. lu a longitudinal direction, tlic lumina, or pores are open. Ileie each 

 lumen acts as a capillary tube, and once moist is as eaj^er to take up water as a capillar\- tube df 

 the same dimensions in plaster of Paris. In pine wood the greatest length of these tubes is about 

 6-7mm. Assuming that the cells are cut in such a man ner t hat t lie average length of the tube exposed 

 by the section is one- half of the total length, or .!-.") nmi., also that the capillary attraction is such 

 that it equals a pressure of one half atmosphere, then the aii- contained in each cell will be reduced 

 to three-fourths of its original volume, allowing one-fourth of the tube to be tilled with watci-. In 

 the case assxmied then, water would till the tubes for about '{-l ram., and if these tubes occnjiy 

 three-fourths of the entire cross section (the other one-fourth being tube, or cell wall) the water at 

 each cross section crowded into the wood, amounts to a sheet less than 1 mm. thick and ■{ of the 

 cross section in extent. Actually tliis maximum is probably never attained. 



Eeturning to the experiments, the above conclusions were largely veritied, but not all. 



To determine the behavior of wood as regards swelling, a jiiece of white pine was caiefidly 

 dressed, measured with a micrometer caliper of Darling, Browne and Sharpe, VJ inch, with vernier 



accurate to one-fiftieth of a millimeter^ ^^^^ inch ) 



The width measured was 4.528 cm. The piece was then measured by immersion, wiped, and 

 measured with the caliper, which had been left set. The caliper still fitted at the particular place 

 which had been marked with pencil. The piece was then iiiiniersed for one minute; the caliper 

 still fitted so iierfectly that no ditlerence in friction on sliding it up and down coulil be obseived. 

 The same was true after an additional inunersion of two minutes and then one of live minutes. 

 The piece of wood was then left todry and was again measured forty-live minutes later. The width 

 had changeil from 4..")28 to 4..532 cm. during the titty minutes since the time of tirst immersion. 



This simple experiment shows what might have been expected, that the imbibition by the cell- 

 walls, on which swelhng depends, is too slow a process ever to interfere in measurements. 



To study the rush of water into the wood, numerous weighings on ditt'erent woods, including 

 pine, both hard and soft, white wood, and oak were made. 



The agreement in their behavior is such that the following illustrations may serve the jiurpose 

 of showing the ettects. A piece of dressed white pine was weighed when dry, then it was meas- 

 ured by immersion, not to obtain the volume but to make the time of immersion that commonly 

 requisite to do the work. The results stand thus : 



Grams. 



Weight when ilry - *^--2 



After first immersion - ^'■^■^ 



After second immersion 85.7 



After tliird inmiirsion 86.2 



After fourth immersion 8().;) 



After fiftli immersion 86.(j 



After sixth immersion 86.6 



After seventh immersion 86.8 



After eighth immersion 87.1 



After ninth immersion 87.1 



After each immersion the piece of wood w^as wiped with a damp cloth, a process whicli re- 

 moves the drippings, but also aids in iierfectly nioistening the wood. Of two |iieces of white w<iod 

 molding, one was given a coat of linseed oil, the other left uiioiled; the foHowing ligur<'s describe 

 their behavior : 



WeiEht. 



Dry , 



After iminiirsion ana wiping 



Oiled piece. 



112. 1 



Xot riiled. 



1 1 2. .I 

 11.-.. 2 



