TRANSPIRATION - Plant physiology · THE COHESION THEORYOF TRANSPIRATION ROBERT E. HUNGATE (WITH TWOFIGURES) The present explanation of the mechanism of water transport in plants (11,

THE COHESION THEORY OF TRANSPIRATION

ROBERT E. HUNGATE

(WITH TWO FIGURES )

The present explanation of the mechanism of water transport in plants(11, p. 9; 12, p. 231) includes, briefly, the pulling of continuous columns ofwater from the soil by means of evaporation from the leaves, the columnsremaining, intact under the stress of great height and rapid evaporationbecause of the tensile strength of the water (6, 16, 20).

WALTER (23, p. 67), after a discussion of experiments on water trans-port, says, "Damit scheint die Theorie der Wasserleitung, die auf derKontinuitit der Wasserfiden von der Spitze bis zur Basis der Pflanzenberuht und deshalb als Kohasionstheorie bezeichnet wird, endgultig gefes-tigt zu sein, wenni auch im Einzelnen immer noch Veranderungen eintretenk6nnen. . . ...

WNTOODHOUSE (24) reports experiments interpreted as indicating a needfor modification of the tensile column theory. The experiments are: first,an inability, using Pittosporium undutlatturt and Ricimis co)nnuntis, to drawmercury to a height greater than 49 cm., and, second, a greater speed ofevaporation from a branch of Ricinus on a day with moderate conditions ofevaporation than on a day of intense evaporation.

In the first case the inability to draw mercury to a height greater than49 cm. can be explainied as due to failure to clean and prepare the apparatussufficiently and to remove air from the plant stem. BOEH-M (4, 5) foundthat out of hundreds of experiments to demonstrate the rise of mercuryabove atmospheric pressure only a few were successful. With more refinednmethods TIIUT (18) and URSPRUNG (21) demonstrated the tensile strengthof a column of water using, living material. THUT (17) a1d OTIS (13) havedescribed methods for its demonstration in a physical system. The experi-ments are sufficiently reliable to use in classroom exercises.

In the second case the difference in rate of evaporation of branches ofcastor bean on the warm and on the cool day canl be ascribed to a probabledifference in leaf area since this factor apparently was not evaluated. Noarguments are presented to support the claim that the development ofgreater tension on a day of less evaporation makes the tensile column theoryinadequate.

In addition WOODHOUSE iinterprets experiments w-ith a 14-meter glassU-tube as throwing doubt on the tensile column theory. "The top of onearm of the U-tube was connected by a stopcock to a vacuum system. It washoped that the velocity of the water, deseendingc, in one arm of the tube and

,83

www.plantphysiol.orgon March 30, 2020 - Published by Downloaded from Copyright © 1934 American Society of Plant Biologists. All rights reserved.

http://www.plantphysiol.org

PLANT PHYSIOLOGY

rising in the other, combined with its cohesive properties, would carry it toan elevation greater than the barometric column. Unlike ASKENASY (3)[1] and THUT (102) [17] under the conditions of this experiment no ten-sile column was obtained. The water entered the vacuum chamber inspurts, and numerous bubbles formed in the column. "

In connection with this experiment it is important to consider the con-dition at the top of any column of water in which its cohesive properties areto be demonstrated. According to RENNER (14): "Die H6he der als un-zerreissbar angenommen Wassersiiule, die in einer beliebig weiten Rohrean einer por6sen Membran aufgehangt werden kann, hiingt von der Poren-weite dieser Membran ab. Die Wassersiiule kann genau so hoch werden,wie das Wasser in einer Kapillarr6hre steigen wiirde, deren Weite auf ihrerganzen LIunge gleich der Weite der gr6bsten Poren der Membran ist. "

In the experiment of WOODHOUSE the diameter at the top of the glasstube in contact with the gas in the vacuum chamber was 1.75 mm. The topof the water column was not attached in any way and could not be sup-ported by its tensile strength. The height to which water rises in such atube is determined by the difference in gas pressure between the base andtop of the column and the force of surface tension at the top. The surfacetension in a tube of the size used will support a column of water to a heightof 18 mm. as calculated from the formula h = 2 a/rdg. The experimentdemonstrates a water barometer. The bubbles appearing in the water whenthe vacuum was applied were very likely due to expansion of air bubblespresent in the joints sealed with de Khotinsky cement.

WOODHOUSE reports a modification of the experiment- in which a threadwas run through the entire tube. In this case a continuous column was ob-tained unbroken by air bubbles. Presumably this column was supportedby additional air pressure at the base since the author states that no tensilecolumn was obtained yet shows a height of 14 meters in the figure (p. 188,24). The rate of flow (in a tube containing water and with a thread fromtop to bottom) when air pressure at the base (i.e., vacuum at the top plusadditional pressure at the base) is released and the tube allowed to emptyis used as an argument for a combined tensile strength-sorption hypothesis.The curve obtained deserves attention. Extension across the abscissa indi-cates that with negative heights the velocity of descent continues to increasein contradiction to accepted laws of hydrostatics and free energy. It isdifficult from the description of the experiment to explain this curve in anyway. It is possible that the de Khotinsky cement extending into the lumenof the tube at certain points may have caused the results.

In modifying the tensile column theory, WOODHOUSE calls sorption intoconsideration, a factor whose importance in water transport in plants hasbeen evaluated by SHULL (15). WOODHOUSE extends the sphere of influence

784



HUNGATE: COHESION THEORY

of the imbibitional or adsorptive forces of the wall material to the center ofthe smaller xylem vessels, suggestingf a limiting distance of 5 microns forthe action of this force. BODE (3) is quoted, " 'Dagegen war das Ein-dringen von Luft in Gefiisse mit einem Durchmesser unter 10 p nie zu beo-bachten.' " The heading of this portion of BODE'S paper is "2. Die Was-serfiiden in mechanisch verletzten Gefiissen." In an earlier portion of hispaper under the heading "1. Im intakten Gefass" BODE states, "Ergebnis:Die Pflanze war vollkommen schlaff, und dennoch bliebein samtliche Wasser-fiiden intakt! !" This was the result after working on Elatostemma sessilewhich has vessels as large as 50 p in diameter.

Aside from the fact that the hypothesis proposed by WOODHOUSE doesnot account for active conduction in vessels exceeding 10 p in diameter,sorption, if acting to the distance and with the force postulated, is an im-pediment rather than an aid to water movement since imbibition forcessufficient to hold water against the pull of gravity would act as effectivelyalong the tube against the pull of transpiration. In his hypothesis the ten-sion of the water is suggested as the motive force while imbibition on thewall is one of support only. But, accordingly, to cause water movementthe tensile force must overcome that of imbibition as well as that of gravitybefore motion can occur. If the imbibition forces are effective to the dis-tance assumed by WOODHOUSE they should be sufficient to prevent the push-ing of water through the 10 p xylem vessels of a short piece of branch care-fully cut at both ends to which a force of 1 atmosphere is applied. DIXON(6) obtained flow of water through wood of Taxuts baccata using a head ofwater equal to the length of the branch. The diameters of the tracheids ofthis plant were not stated but presumably they were not much larger than10 p, while the diameter of the pores in the walls of the tracheids did notexceed 0.3 p.

ExperimentsTo establish a complete water column supported by its adsorptive and

cohesive forces the apparatus shown in figure 1 was set up. The tubes ABand CD were formed by melting with a gas-oxygen flame two Pyrex tubes60 by 1.5 cm. and gradually drawing them out in a stair well into a tube ofvarying size, about 1.5 to 6 mm. inside diameter. The final length was 13.5meters. After both tubes were drawn they were sealed together at the topso that a continuous glass siphon was formed. In order to insure sorptionof the water on the glass, potassium bichromate in sulphuric acid was forcedfrom D to A by means of a tire pump and valve at F. The flask used was a250-ce. round-bottomed Pyrex. flask. The rubber stoppers were wired inplace to hold the pressure which was estimated by means of the manometerG. To obtain fairly rapid passage of the solution it was necessary to employpressures of as much as five atmospheres. NVhen the tube had been filled

785



76PLANT PHYSIOLOGY

A ~

FIG. 1. Apparatus to demonstrate cohesion of water column.

with the cleaning solution alcohol was poured down the outside from thetop and ignited when it reached the bottom. This heating of the solution inthe tube insured thorough cleaning.

1. COHESION OF A COLUMN OF MOVING AIR-FREE WATER

The cleaning solution in the tube was followed by boiling water and thisby boiled water which was allowed to cool in the tubes. The pressure in theflask was released by unscrewing the clamp H. The water column did notbreak and the 14-inch difference in level between A and E caused water todrip from A. After about 30 cc. of water left in the flask were emptied theascending meniscus moved up the shorter arm DC with increasing speed asthe difference in water level increased.

786



HUNIGATE: COHESION THEORY7

2. COIIESION OF A COLUAIN OF MOVING AIR-CONTAINING WATERThe experiment was perfornmed as before except that unboiled distilled

water was substituted for the boiled water. This distilled water had beenstanding in a 5-gallon bottle open to the air. In this experiment the watercolumn remained intact when pressure was released at the base and siphonedlfor six minutes, emptying 20 cc. of water at the base of the lower arm. Thecolumn then broke ten feet from the top of the tube DC, probably due tothe conduction above 32 feet of somne impurity in the distilled water. Tlleflow before breaking w-as sufficient to show that the cohesive force of a mov-ing solution of air in water is strong enough to support the water columni toa height of 44 feet.

3. FORCE OF WATER SORPTION IN AN AGAR GEL

Tube AB was disconnected from CD at the top of the column and onieend of a short piece of glass tubing 3 cm. in diameter was sealed on to theopen top end. The junction of the small and the 3-cm. tubing was then niar-rowed to an area of about 3 sq. mm. The top of the 3-cm. tubing was leftopen to the atmosphere. Boiled water was forced up the tube after clean-ing with cleaning solution and allowed to cool. The pressure at the basewas maintained so that the water colunil was supported but did not clhanoein level. Two per cent. agar solutioni was prepared and filtered while hotthrough a fritted glass filter. A piece of absorbent cotton was then dippedin the agar, taking care that no air bubbles were included in the intersticesof the cotton which was tllen placed in the top of the water column andforced down into the conistriction. The excess water in the top portioll ofthe tube was replaced by agar and tile cotton and agar allowed to set. Thecolumn remained intact when the pressure at the base was released.

It was ori(rinally hoped that evaporation from the surface of the agarwould result in drawing water up the tube. To this end a waxed papercylinder was tied over the open end of the tube containing the agar andconnected to a beaker of sulphurie acid. The agar gradually decreased involume and shrank away from the sides of the tube until at the end of 78hours the shrinking allowed air to slip past the cotton-agar plug and thecolumn broke. Only a part of the water leaving the agar gel was absorbedby the sulphuric acid, since 2 cc. of water were collected in a 10-cc. gradui-ated cylinder placed at the bottoim open end of the tube to measure thevolume change. That this was not a volume change due to temperature dif-ference was evidenced by tile fact that the increase in water at the bottomwas gradual during the time that the column was intact. The tensioni (lessthan 1 atmosphere) at this height was sufficient to remove some of the waterfrom the gel. The imbibition forces in the gel are insufficient to hold all thewater againist this pull. If the experiinent hlad continued until equilibrium

r' 8



PLANT PHYSIOLOGY

had been established and no more water withdrawn, presumably evapora-tion from the surface of the agar would have resulted in pulling water upthe tube.

Under such conditions of equilibrium the water bouind in the gel by a forceof about 0.5 atmosphere could be determined by drying the agar and weigh-ing the residual water. Although at completion of the experiment equilib-rium had not been reached, a considerable part of the water in the gel waswithdrawn as is indicated by the figures (table I) for the 78-hour period.

TABLE I

Agar in the original gel .................................................................................................................................. 2 per cent.Weight of agar, water, and cotton after 78 hours.6.40 gm.

" " " andcotton.0.46 gin." " cotton after boiling in water and drying.0.28 gm." " agar.0.18 gn." " water.5.94 gn.

Percentage agar after 78 hours of 0.5 atm. tension ............................................................. 2.94 per cent.Percentage of water removed from agar during tension .... ...... 32.7 per cent.

Of the water lost from the agar, 2 cc. were collected in the graduatedcylinder at the bottom. The rest was absorbed in the sulphuric acid at thetop of the column.

4. COHESION OF A STATIONARY COLUMN OF AIR-CONTAINING WATER

It was believed that difficulties in demonstrating the cohesive force ofair-containing water were not so much due to dissolved air as to the pres-ence of small particulate impurities with adsorbed air which might providea nucleus for collection of dissolved air (6). The importance of such nucleican be qualitatively observed in a dirty beaker in which tap water is heatedas compared with a beaker cleaned in very hot cleaning solution. In thefirst case bubbles gradually collect on the walls of the beaker as the tem-perature rises, whereas in the second case few bubbles a-re to be observed.The influence of dirt on the boiling point was noted by GAY-LUSSAC [afterURSPRUNG (19)]. Observations similar in principle indicating the impor-tance of particles in allowing condensation of water vapor from the air areknown from meteorological observations as well as from experiments inphotographing electron paths.

Instead of boiling the water the results of experiment 3 suggested thepossibility of filtering out all particulate impurities by means of the agar.A tube similar to that of experiment. 3 was prepared except that it wasdrawn quickly from Pyrex tubing 3 cm. in diameter, forming a capillary13.9 m. long, of diameter 0.6 mm. in the center to 1.4 mm. in diameter at a

788




distance 1 m. from the end. It was cleaned, filled with boiled, cooled water,and plugged with a cotton-agar plug. After the gel had set, the pressure atthe base was released with no resultant break in the water column. Eight cc.of 0.25 per cent. acid fuchsin solution were poured on top of the agar. The topwas closed to evaporation by tying over it a waxed paper bag. The layerof agar and cotton was about 2.5 cm. in thickness, tapering from a diam-eter of 3 cm. at the top to one of 1.5 mm. where the capillary joined thelarge tube. Below the constriction in the water-filled portion of the tubethe diameter was 6 mm. In 8 days the acid fuchsin solution on top of theagar had been sucked through and an equivalent amount of water had col-lected in the flask at the bottom of the capillary. The red color of the dyecould be observed extending from the agar down the capillary to a distanceof 1 meter. Unless by some means the air in the dye solution was removedduring the passage through the agar the water at the top of the column nowcontained dissolved air. It did not break.

To obtain a greater suction tension the container at the base of the capil-lary was gradually evacuated to determine at what tensile stress the columnwould break. When the manometer indicated 35.3 cm. Hg the sudden in-crease of water in the flask indicated that the column had broken. Thebreak occurred at the junction of the agar-cotton gel with the water column.The barometric pressure was 75.3 cm. Hg. Calculation shows that thebreak occurred under a tension of 0.9 atmospheres. Thus a column of air-containing water is supported by its cohesive force to a height of at least 9meters (19 meters when open to the atmosphere).

5. EFFECT OF TENSION ON VISCOSITY OF WATER IN A CAPILLARY TUBE

In connection with the speeds of water movement found in plants (8), itis of interest to determine whether viscosity in a water column decreaseswith tension. Two capillary tubes were drawn out to a distance of about 35feet in the same way as in experiment 4. These were sealed together at thetop and at the base were sealed into flasks as indicated in figure 2. Withthe apparatus as shown it was possible to apply pressure or suction at willto either or both of the flasks at the bases of the capillary tubes and to countvery accurately the flow of water by means of the drops from the capillaryin the lower flask. After cleaning and filling with boiled distilled waterthe flasks and tubes were allowed to cool and then to siphon for two hours.By this time a temperature equilibrium with the surroundings had beenreached.

Measurements were taken: (1) when the water in the flasks at the basewas subjected to atmospheric pressure; (2) with an additional pressure of40 cm. Hg applied to both of the flasks; (3) at atmospheric pressure as in

789



()PLANT PIHYSIOLOGY

FIG. 2. Apparatus for measuring viscosity of water under tensioni.

1; and (4) with the flasks at the base both evacuated to a pressure of 27 cm.Hg. The results are given in table II.

Changes in viscosity with this tension change in the water are not suffi-cient to be detected with the nmethod of measurement used. An experimentwas started to determine the above rates with the use of a stopwatch butwas discontinued when one of the flasks burst under the pressure used infilling the tube.

It is apparent that the method of measuring the siphoned water dependsupon its surface tension as it drops from the lower capillary. In case the

TABLE IIMEASUREMENTS OF THE VISCOSITY OF WATER AT DIFFERENT TENSIONS

PRESSURE TIME No. DROPS RATE

76 cm. Hg 3 min. 55 sec. 20 11.75 drops/sec.116 cm. Hg 3 min. 54 sec. 20 11.7 drops/see.76 cIll. Hg 1 min. 57 sec. 10 11.7 drops/sec.76 cim. Hg 1 min. 57 sec. 10 11.7 drops/see.2 7 cm. Hg 1 min. 56 sec. 10 11.6 drops/sec.

79()

I JL..j

/1"',ll-'-ll'll.."l'l'.-1-



HUNGATE: COHESION TIIEORY7

surface tension of the water changes in exactly the same way as the viscositywhen subjected to varyina tension or pressure the above experiment is in-validated. It would be necessary in such case to substitute some othermeans of measuring the water.

Discussion

To obtain consistent results in experiments in which the forces of adsorp-tion and cohesion are concerned it is important to work with materials andsurfaces that are clean and homogeneous. The variations in the experi-mental results obtained by various workers who have attempted to measurethe cohesive forces of water are probably due to the variations in the degreeto which they succeeded in cleaningf the apparatus with which they worked.

In connection with impurities in the water it is important to distinguishbetween the state of different kinds of impurities. Gases, solids, and liquidsdissolved in water are present in dispersed molecular or few molecularunits. Any tendency of these impurities to lessen the cohesive force of thewater is spread evenly throughout the mass of the water and not concenl-trated at a surface. (When concentration of the solute occurs at the surfaceas in soap solutions there is a very great effect on the surface tension andperhaps a corresponiding effect on the cohesive force.) Some substances insolution may increase the cohesive force of the water.

The cell wall absorbs water with great force, insuring thorough wettingof the surface with wMhich the water is in contact. That imbibingr colloidssuch as the cell wall need niot act to prevent movements of water throughthem whlen subjected to tension is indicated by the passage of the waterthrough the agar oel. This may explain in part the way in which thetranspiration stream travels from cell to cell ini the w-ater-conducting sys-tems of some plants.

That a gel wlhen subjected to a suction tension nmay shrink while stillmaintainingr its orgyanization anid structure intact is indicated by the be-havior of the agar oel. The relatively low suctioni tension force used inextractilng +water from the gel slhows that the force of water sorption is lowfor a conisiderable proportioni of the water held in the gel. Sinice the agargel is a colloid the solid matter composing it may be assumed to be dis-persed in units of about 5-200 mp in diameter. It seems inot unreasonableto assume that the dliameter of the -water-filled spaces in the solid phasedoes niot exceed 200 inp. The spaces are not visible microscopically. If thewater spaces in the gel are of this size it muist be conieluded that the forcesof water sorption of any magniitude are confinied to a very short distance inthe colloid since with a relatively low suction tension a considerable amountof this water can be removed. It seems probable that in the water-coniduct-ing cells of the planit great forces of water sorption act for very short dis-

791



PLANT PHYSIOLOGY

tances and insure a wall which will act efficiently in excluding air bubblesand in providing an easily wetted surface, but they do not entirely impedethe movement of water in the liquid phase within the colloid, much less inthe lumen of conducting cells.

The results obtained in experiment 3 suggest that by using tensions ofwater a quantitative measure of the amounts of water held in certain col-loids can be determined as well as the forces with which they are held. Itmay be that some modification of the scheme could be used to determine theamounts and forces concerned in the "binding" of water in non-living andperhaps in living systems.

The variations in trunk diameter of trees (10, 7) seem quite reasonable ifwe suppose that the tension of the water during periods of rapid transpira-tion draws a certain small amount of water out of the colloidal material ofwhich the cell wall is composed. When the tension is released the im-bibitional forces of the cell wall cause an absorption of water into the in-terstices of the colloid composing the wall and an increase in the size of thecell. It is of interest to speculate as to what effect the shrinkage of thecellulose in the wall might have on the ease with which water columns underheavy tension might be broken. It would seem that the shrinkage shoulddecrease the ease with which air might penetrate into the vessels. Anothereffect of wall shrinkage may be to increase the rigidity of cells subjected toa tension so that leaf cells under normal tensions do not wrinkle and col-lapse but remain in approximately the same shape as when distended byturgor.

URSPRUNG and BLUM (22) have measured suction tensions in variousparts of the plant and shown that the forces found are in accord withthose needed in a cohesion theory of sap transport. To the writer thistheory provides a coherent picture of the hydraulics of transpiration, apicture supported by many diverse experiments all indicating the simplic-ity and clarity of the central idea.

If we accept the physical forces operating at the surface of the leaf asthe moving factors in supplying water to the plant, the constant wiltingcoefficient of a given soil for many kinds of plants finds a possible explana-tion. Evaporation from the mesophyll cells of the leaves draws ultinlatelyupon the water in the soil. Since the walls of the evaporating cells ofplants are composed of substances physically similar, celluloses, it is reason-able to conclude that they will have a similar maximum force which theycan exert upon the water in the soil. Wilting occurs when the force withwhich water is held in the soil exceeds the force which the evaporating sur-face of the leaf can exert. Use of an instrument absorbing water throughphysical forces (9) has given a fairly constant value for the wilting pointsof various soils, a value which, if expressed in absolute amounts, is quitevariable.

792




Summary1. The results of certain experiments interpreted as throwing doubt on

the tensile column theory of the ascent of sap can be explained as due toinadequate technique.

2. The tension-sorption hypothesis to explain ascent of sap duringactive transpiration is experimentally unsupported.

3. Reports of cohesive forces in air-free and air-containing water aresubstantiated.

4. A method is indicated by which bound water may be determined.5. Within the limits of error of 1 per cent. no change in viscosity with a

tension change of 1 atmosphere was found although the method of measur-ing the water may have hidden such changes.

6. A mechanism is suggested by which size changes without formchanges may occur in both hard and soft parts of plants.

STANFORD UNIVERSITY

LITERATURE CITED1. ASKENASY, E. Uber das Saftsteigen. Verh. naturhist.-med. Ver. Hei-

delberg N.F. 5: 325-345. 1895. Cited by WOODHOUSE.2. BACHMANN, F. Studien fiber Dickenainderungen von Laubbliitern.

Jahrb. wiss. Bot. 61: 372-429. 1922.3. BODE, H. R. Beitrage zur Dynamik der Wasserbewegung in den Ge-

fasspflanzen. Jahrb. wiss. Bot. 62: 92-127. 1923.4. BOEHM, J. Ursache des Saftsteigens. Ber. deutsch. bot. Ges. 7: (46)-

(56). 1889.a. . Capillaritait und Saftsteigen. Ber. deutsch. bot. Ges.

11: 203-212. 1893.6. DIxoN, H. H. Transpiration and the ascent of sap in plants. Lon-

don. 1914.7. HAASIS, F. W. Seasonal shrinkage of MNIonterey pine and redwood

trees. Plant Physiol. 7: 285-295. 1932.8. HARVEY, R. B. Tracing the transpiration stream witlh dyes. Amer.

Jour. Bot. 17: 657-661. 1930.9. LIVINGSTON, B. E., and KOKETSU, R. The water-supplying power of

the soil as related to the wilting of plants. Soil Science 9: 469-485. 1920.

10. MIAcDOUGAL, D. T. Growth in trees. Carnegie Inst. of Washington.Pub. no. 307. 1921.

11. , OVERTON, J. B., and SMITH, G. M. The hydrostatic-pneumatic system of certain trees: movements of liquids and gases.Carnegie Inst. of Washington. Pub. no. 397. 1929.

793"



PLANT PHYSIOLOGY

12. MAXIMOV, N. A. The plant in relation to water. Translation byYAPP. London. 1919.

13. OTIS, CHARLES H. The ASKENASY demonstration. Plant Physiol. 5:419-423. 1930.

14. RENNER, 0. Die Porenweite der Zellhiiute in ihrer Beziehunog zumSaftsteigen. Ber. deutsch. bot. Ges. 43: 207-211. 1925.

15. SHULL, CHARLES A. Imbibition in relation to absorption and trans-portation of water in plants.- Ecology 5: 230-240. 1924.

16. SMITH, FANNY, DUSTMAN, R. B., and SHULL, C. A. Ascent of sap inplants. Bot. Gaz. 91: 395-410. 1931.

17. THUT, H. F. Demonstration of the lifting power of evaporation.Ohio Jour. Sci. 28: 292-298. 1928.

18. Demonstrating the lifting power of transpiration.Amer. Jour. Bot. 19: 358-364. 1932.

19. URSPRUNG, A. Zur Demonstration der FlUssigkeits-Kohiision. Ber.d-eutsch. bot. Ges. 31: 388-400. 1913.

20. . Uber die Kohiision des Wassers im Farnannulus. Ber.deutsch. bot. Ges. 33: 153-162. 1915.

21. . Dritter Beitrag zur Demonstration der Fliissigkeits-kohasion. Ber. deutsch. bot. Ges. 34: 475-488. 1916.

22. , and BLUM, G. Eine Methode zur Messung polarerSaugkraftdifferenzen. Jahrb. wiss. Bot. 65: 1-27. 1925.

23. WALTER, HEINRICH. Der Wasserhaushalt der Pflanze in quantitativerBetrachtung. Freizing-Miinchen. 1925.

24. WOODHOUSE, E. D. Sap hydraulics. Plant Physiol. 8: 177-202.1933.

79-M



Documents

TRANSPIRATION - Plant physiology · THE COHESION THEORYOF TRANSPIRATION ROBERT E. HUNGATE (WITH TWOFIGURES) The present explanation of the mechanism of water transport in plants (11,