116 



SCIENCE. 



[N. S. Vol. XX. No. 499. 



mental impression too faint for complete iden- 

 tification, now that attention is directed to it, 

 nevertheless rose into consciousness with the 

 semblance of a spontaneous idea, and gave 

 rise to a distinctly plagiaristic publication. 



G. K. Gilbert. 

 San Fbancisco, 

 June 28, 1904. 



SPECIAL ARTICLES. 



THE ASCEKT OF WATER IN TREES. 



Recent discoveries by Vesque and E. B. 

 Copeland and others have brought us very 

 near the solution of this inveterate problem; 

 but botanists seem agreed to halt at the last 



f' 



'Area a\ 



;Cross-5ec.l 



'4s<T. cm^ 



Hijdrostatic 

 Pa. ad 



ox. 



step, awaiting some occult signal from the 

 physicists. The old toy of the ' hydrostatic 

 paradox ' ought to teach them that water 

 pressed upwards by the atmosphere has no 

 divine right to call a halt at 1,033 centimeters. 

 One fluid may support and also elevate an- 

 other fluid to any required height. Thus the 

 supported weight in the annexed diagram may 

 be represented by a column of water raised 

 a mile high or more. The condition is 

 that there shall be no immediate continuity 



of mass between the fluid to which the 

 atmospheric pressure is applied, and that 

 which is to be lifted. This condition is se- 

 cured in the tree by the numerous transverse 

 septa on its water-ducts, which prevent the 

 transmission of air or water in mass, but per- 

 mit a very free molecular diffusion of water, 

 and of everything dissolved in it. 



Assuming the diameter of a water-duct to 

 be half a millimeter, it is easy to estimate the 

 weight supportable by a septum at its base; 

 seeing that its upward parts are protected 

 from other atmospheric pressure. The area 

 of the cross-section of the basal part, in centi- 

 meters, if multiplied by 1,033 (the height in 

 centimeters of a column of water equaling 

 the atmospheric pressure), gives 2.028 cubic 

 centimeters, or the sam.e number of grams, 

 as the load which the basal septum can sup- 

 port. This load may be a column of water 

 filling the whole duct for a height of 34 feet, 

 or it might be a column or stream of water 

 twelve times as high and only- one twelfth as 

 large in cross-section. If the duct were filled 

 with one part of water and eleven parts of air, 

 the water and air together should be supported 

 by the atniospheric pressure at the base to a 

 height of about 403 feet. Assuming that 

 there were in this course about 90 cross-septa, 

 approximately equidistant, we should have a 

 fall of pressure with each succeeding stage, 

 equivalent to about one third of an inch of the 

 mercurial barometer, reaching zero at the 

 summit. 



The mechanism in the xylem-ducts, however, 

 can not be of this kind ; because not only would 

 the gas-bubbles obstruct the current if they 

 came between it and the walls, but the spread- 

 ing out of the pressure of the ascending cur- 

 rent over the septa would cause the ' hydro- 

 static paradox ' to work backwards with greatly 

 increased force. This compels us to favor the 

 view of interning the gas-bubbles within the 

 water. Dr. MacDougal states that ' the 

 cavity of a wood-cell contains a bubble of 

 gas' C Plant-Physiology,' p. 29). And Stras- 

 burger describes the water as freely streaming 

 round the gas bubbles, or between them and 

 the walls. This arrangement of water sur- 

 rounding gas-bubbles constitutes what is 



