490 ANNUAL REPORT SMITHSONIAN INSTITUTION, 19 3 8 



been accepted not so much because of its freedom from objections 

 as because of the inadequacy of all other theories. The only real con- 

 tender — Hales' old "root-pressure" scheme — was rejected, first, be- 

 cause pressures greater than the 1.4 atmospheres recorded by Hales 

 himself had not been observed, and, second, because all demonstra- 

 tions of root-pressure rested on experiments with decapitated and 

 moribund plants subject to Molisch's criticism. Water must be 

 raised in some Eucalyptus trees to a height of 350 feet, requiring either 

 a push or a pull of at least 13 atmospheres. The maximum root-pres- 

 sures observed were only of a magnitude of 1 .4 atmospheres, they were 

 transitory, and doubt even existed as to their presence in uninjured 

 plants. Consequently, modern plant physiology texts for the most 

 part treat "root-pressure" as unimportant. 



Excised roots of tomato have been kept growing in vitro for a num- 

 ber of years (White, 1937). The roots form normally developed 

 vascular strands, although completely immersed in a nutrient solution 

 (pi. 1, fig. 1). The occurrence of strands in roots grown under these 

 conditions has seemed rather anomalous. If, however, a continuous 

 flow of liquid be assumed to take place through the roots, the presence 

 of strands would be understandable. If such a current exists, it was 

 thought these roots might offer a meaDS of determining the reality or 

 unreality of "root-pressure" and perhaps of measuring it under con- 

 ditions free from Molisch's objections. 



Means have, therefore, been devised of repeating Hales' original 

 experiment, using single actively metabolizing tomato roots instead 

 of his moribund grapevine stocks. 



Capillary manometers were built to receive single roots. Specimens 

 of a clone of excised tomato roots grown in continuous culture for the 

 past 5 years were subcultured and allowed to stand for 1 week, to 

 give the cut surfaces time to heal completely. Their bases were then 

 carefully inserted into manometers, and seals were made by means of 

 miniature rubber hose corresponding to those used in the classic 

 Fuchsia demonstration known to every student of plant physiology. 

 The roots with their attached manometers were returned to fresh flasks 

 of nutrient, and their subsequent behavior observed (pi. 1, fig. 2). 

 All manipulations had, of course, to be carried out aseptically and 

 with as little trauma as possible. 



The results of the first experiments were surprisingly good (White, 

 1936). The roots did secrete water from their bases into manometers. 

 There does exist a unidirectional flow of liquid through these roots. 

 "Root-pressure" is not an artifact but a reality. This paper proposes 

 to present some quantitative results of these experiments. 



If roots of this land are set up in two series of manometers, with 

 capillaries of the same diameter, one series containing water, the other 

 mercury, we would expect a decrease in the secretion rate under the 



