p. Groom 



29 



During shrinkage the board did not dry completely to its original 

 condition (it retained onlyabout 5-5 g. more water), but making allowance 

 for this fact, its shrinkage was not relatively (nor absolutely) so great as 

 the swelling. The percentage average decreases in linear dimensions 

 were : in length -063 ; in width at the ends and middle 5-92 and 5'98 

 respectively. The percentage tangential shortening was 94 times as 

 great as the longitudinal. 



Changes in area. During absorption and emission of water it has 

 been shown that synchronous longitudinal and tangential changes in 

 dimensions may be opposite in sense. The question arises therefore: 

 During absorption and emission of water does the volume or surface 

 as a whole always respectively increase and decrease? The statistics 

 available prove that a decrease and increase in the amount of water 

 respectively cause corresponding decrease and increase in area : thus 

 anomalous changes in length are overborne by the more normal changes 

 in width (the one apparent exception recorded in Table C ii July 19th 

 is discussed later). 



The subjoined Table D records the changes in surface for each 

 ten grammes of water gained or lost. In order to show the corresponding 

 conditions as regards the statistics concerning drying and soaking 

 respectively the data are ranged in columns in reverse order. 



Both sets of observations show that the greatest change in area (and, 

 judging by the statistics concerning the radial board, in volume) induced 



