440 



THE AMERICAN NATURALIST [Vol. XL 



vessels, and in some of the wood fibers. The starch-containing 

 fibers were mostly either adjacent to the vessels or clustered in a 

 band at the end of each year's growth. The other wood fibers 

 contained none at all. All of these cells with starch are living 

 and contain protoplasm. In autumn we find the starch-con- 

 taining cells of the maple packed full of this substance as in other 

 trees. From the time cold weather commences until springs 

 starch is gradually converted into sugar. Fischer ^ has found 

 this to be the case in many trees. The Vermont workers also 

 found the starch content to decrease in early spring, and the 

 sugar-content to increase. There is, however, no evidence to 

 show that sugar is again reconverted into starch in late spring as 

 Fischer states to be the case in some trees. It seems that the 

 starch stored in the pith-ray cells and in the wood fibers described 

 above is gradually converted into sugar as spring advances. Since 

 there is no other source for the constantly increasing sugar content 

 of the sap in the vessels it seems reasonably certain that this sugar 

 escapes into the vessels from the starch cells where it is formed. 



If pressure and flow are due to the living cells they must then 

 be due to the pith-ray system, the wood-fiber system, or both, since 

 these constitute the living part of the wood. In wood, as shown 

 by Sachs, the only direction in which water passes with dilliculty 

 is radially. In longitudinal and tangential directions tlwrc is 

 little obstruction to the flow. Moreover, althmi-h in the previous 

 discussions in this paper, the fiber walls were li\ jxiilietically con- 

 sidered as almost impermeable to water, it Mcin> morr likely 

 that one or even two walls intervening would retard the j a^aue 

 of water but little. At any rate, it is scarcely probable that a 

 strand of starch-containing wood fibers contiguous to a vessel at 

 one end would be more than two or three walls distant at the other 

 end, and therefore there would be no way of obtaining for such a 

 strand the two distinct water reservoirs neees-ary for the j)r()- 

 duction of pressure. 



In order that a homogeneous niemhrane should heeome suddenly 

 more permeable at certain regions at a definite rising temperature 



'Fischer, A. "Beitriige zur Physiologic der Holzgewachse." Prinqsh. 



