1909} MERWIN AND LYON—SAP PRESSURE IN THE BIRCH 451 
glucose. A sap containing o.8 per cent. of glucose represents an 
osmotic pressure of about 78°™ at o° C. It follows, then, that roo°™ 
of pressure in a birch stem is the maximum to be expected from root 
pressure. 
Volume changes in the sap and wood due to changes of temperature 
in the tree cause marked variations in pressure. 
I find that the expansion of sap from 6° to 32° C. is only 2 per cent. 
greater than the expansion of water. Cell wall substance, on the 
other hand, when saturated with water, expands about 2.2 times as 
much as water between 6° and 32°C. (exp. 3). 
Observations as to the elasticity to the transmission of light of birch 
wood tissue in a thin microscopic section shows that wood fibers 
and the walls of the vessels in the wood have the least elasticity 
parallel to the length of the stem, and that the medullary rays have 
least elasticity along radii of the stem. In a wood fiber the greatest 
elasticity is perpendicular to the surface. By comparison with other 
substances, in which expansion by heat is directly related to elasticity 
to light, a different coefficient of thermal expansion for different 
directions would be expected in both single wood fibers and in masses 
of wood. My determinations made on strips of green white birch 
wood about 500°™ long immersed in water show that between 6° and 
32° C, there is contraction instead of expansion in a radial direction 
when the temperature is raised. Under like conditions a longitudinal 
strip showed at first a slight expansion, but in two subsequent deter- 
minations it contracted. The coefficients of radial contraction 
obtained were 0.000005, 0.000004, and 0.000006; and those of 
Jongitudinal contraction were 0.000002 and 0.000003. These 
coefficients are so extremely small that they may be neglected in the 
following sap pressure calculations.5 It thus appears that the volume 
changes in the cell walls above mentioned are made possible only by 
a diminution of the area of cross-section of the vessels and of the 
cavities in the cells, for the external dimensions of the tree change 
Scarcely at all. 
The effect of this tendency to diminish the pore space in the wood 
’ The thermal expansion—based upon exp. 3 and upon the above coefficients— 
of a given volume of birch wood, of which 4o-45 per cent. is saturated cell wall, amounts 
‘6 about 1.5 times the expansion of an equal volume of water. 
