iv.] VARIOUS THEORIES, &c. 87 



the walls of which are almost impermeable to water. 

 Elfving's experiments prove that no such easy mobility 

 of the imbibed water, as Sachs assumes, exists ; and 

 Hartig confirms the view that the water only moves 

 through the membranes of the bordered pits. 



These delicate closing membranes are very elastic, 

 and when extended by pressure are particularly thin 

 and permeable ; the solid ring-border is a support so 

 arranged that when the delicate membrane is driven 

 too far it rests on the inner surface of the ring, and 

 the torus blocks up the pore, the apparatus thus 

 acting as a safety-valve to prevent undue tension or 

 rupture of the filter-membrane. 



In Dicotyledons water can be more easily forced in 

 a radial direction than in Conifers, because the 

 bordered pits are not confined to the radial walls ; but 

 even in Conifers the last-formed tracheides of each 

 annual ring have numerous very small bordered 

 pits on their tangential walls, no doubt to serve as 

 water-doors to supply the cambium in the spring, as 

 otherwise it must suffer. Finally, it should be noted 

 that in the Conifer the long, prismatic tracheides are 

 arranged in the annual ring in radial rows, those 

 in each radial row being equal in height : tangentially, 

 however, the rows stand at unequal heights, so that 

 anything passing through the bordered pits (on the 



