July 22, 1904.] 



SCIENCE. 



117 



known as foam; not, indeed, of the rough 

 kind, but ' dressed ' so as to be in unison with 

 the shape of the duct through which it must 

 pass. It would take a great many hundreds 

 of meters of such foam to weigh a kilogram to 

 the square centimeter. Thus the atmospheric 

 pressure at the base proves to be sufficient for 

 the work to be done. Every change of equi- 

 librium will cause a movement upwards of the 

 water which is the only movable ingredient 

 of the mixture. 



If we are correct in accepting the observa- 

 tions that the water thus surrounds and en- 

 closes the continuous or beaded air-globules 

 (which must also have much vapor of water) 

 not only is the streaming of the water ac- 

 counted for, but also such phenomena as 

 capillarity and diffusion, and occasional stasis 

 reminding one of the phenomena of capillary 

 circulation of blood. Also the correlation 

 which Strasburger, Vesque and others have 

 observed between the state of the barometer 

 and the streaming within the xylem-ducts is 

 explained; and possibly the pulsation of gas- 

 bubbles which MacDougal regarded as helping 

 to raise the water upwards. 



The condition of low apical pressure is se- 

 cured by the activity of the leaves correlated 

 ■with the structure of the ducts. The leaves 

 are not known to actively attract the water, 

 but they always remove it as it arrives, 

 turning it into starch, and transpiring it in 

 great quantity. When they die or are 

 stripped, the ascent of water ceases (though 

 at lower parts of the stem bleeding may be 

 thereby induced). Also in the leaves and 

 downwards the water-ducts are protected from 

 direct atmospheric pressure by their structure, 

 having spiral threads, and bars and thick walls 

 to prevent compression, and having their very 

 tips roofed over by domes. Their pits with 

 thin membranes permit diffusion sideways, 

 into other ducts or into parenchyma ; and 

 thus interchange is secured among the dif- 

 ferent elements. The machine will not work 

 if the walls are torn, but if they are whole, 

 their being dead does not obstruct their ac- 

 tion. The fact that trees become stag-headed 

 from a loss of water proves a delicate adjust- 

 ment, especially as the different species have 



their differemces of height and other idiosyn- 

 crasies relatively to the phenomenon. 



The transverse septa of the water-ducts ap- 

 pear to us to play a leading part in the process. 

 Transmitting the water by diffusion and inter- 

 cepting the gases, they constitute a series of 

 chambers each having a number of immobile 

 air-bubbles, and permitting the water to flow 

 between these by the only channel which it 

 can take. We do not know how thick is the 

 stream, otherwise than by the deductive 

 method given above, but it probably varies in 

 thickness and rapidity relatively to the height 

 and leafage of the tree; very lofty trees hav- 

 ing little leafage, and yet having many water- 

 duets. We think that the dynamical part of 

 the problem is explainable mainly by the vis 

 a tergo of the atmospheric pressure at the base. 



The tracheids of the giant conifers exhibit 

 the same principle, by their bordered pits with 

 a torus which is centrally thickened and is 

 overarched by diaphragms. The torus is a 

 relatively large flexible membrane, which 

 transmits the whole pressure, and at the same 

 time limits the quantum of water that can find 

 a passage upwards. 



The air drops are not only obstructed by the 

 septa, but they seem to keep apart within a 

 section, as if we had a combination of the 

 Jamin-theory with the osmotic functions of 

 the septa. Each air-drop seems to have a 

 shell which prevents its fusing with its neigh- 

 bors, and thus the system becomes a sort of 

 emulsion, like the fat-globules in milk. My 

 colleague. Professor E. H. Loomis, who has 

 aided me by criticizing the physical points, 

 furnishes me with a striking illustration of 

 this phenomenon. A bubble of air having 

 got into one of his barometers, and being im- 

 prisoned between the mercury and the glass, 

 he let in other bubbles. But these carefully 

 avoided it; and when their course seemed to 

 indicate collision and fupion, they turned 

 aside and passed round it, escaping contact. 

 Another colleague tells me of a case in 

 Kansas where the lives of a community were 

 saved during a drouth by a rotten pump which 

 admitted air as well as water and raised the 

 mixture about 40 feet after the other pumps 

 had given ovit. 



