426 



THE AMERICAN NATURALIST [Vol. XL 



section of the trunk would be tributary to the tap-hole, but only 

 the few outer layers. If this be true, the flow to be expected from 

 this source would be an extremely small fraction of the whole 

 flow, not more than s^g to -5^. It is obvious, therefore, that sap 

 expansion cannot account for the flow in the maple. 



Moreover, it is also not easy to see why air should pass back 

 into the wood fibers on cooling below 0° C. at a time when there is 

 actual suction in the vessels. Even if this were possible the 

 theory cannot account for the flow, and can account for the pres- 

 sure only by supposing the walls of the wood fibers impervious to 

 water to an extent beyond the range of probability. The water- 

 expansion theory must therefore be considered almost, if not 

 quite, out of the question. 



The Wood-expaxsiox Theory 



There is still another possible source of pressure due to heat 

 expansion, namely, the expansion of the wood itself. Wood ex- 

 pands, as well as swells, more in transverse than in longitudinal 

 direction. The coefficient of expansion as determined by Villari ^ 

 for dry maple wood is 0.000,006,38 in longitudinal direction, and 

 0.000,048,4 parallel with the radius. The coefficient for wet wood 

 is not recorded, but it must be considerably greater. 



Let us suppose that the rising sun falls abruptly upon the tree 

 or that the air temperature itself rises rapidly; then the outer 

 layer of wood will be warmed much more rapidly than the inner, 

 and probably the rise will be several degrees in the outermost 

 layers. These outer layers will tend to expand, but being firmly 

 united with the inner, such expansion is possible only to the extent 

 of the elasticity of the wood. The result will be, therefore, a pres- 

 sure among the elements composing the outer layers. The vessels 

 of maple wood are quite large and surrounded by fibers. Since 

 the latter contain the bulk of the wall-substance, they, in expand- 

 ing, would tend to press upon the large cavities, the vessels. Since 



nn. d. Physik u. Chemte, vol. 133, p. 400, 1868. 



