OBSERVATIONS ON THE DAUBREE EXPERIMENT 5 



which may equal, or even exceed, the barometric height/ The 

 mercury rises gradually — ^provided that the pores are not too fine — 

 maintains the equilibrium position for some time, and then drops 

 back rapidly; the rate of rise, which depends upon a number of 

 factors, does not concern us here. The position of equilibrium is 

 the important thing; it depends only upon the size of the widest 

 pores at the free surface and the surface tension between water 

 and the porous material; results for samples of various materials, 

 as observed by us, are brought together in Table I. But before 

 considering these results we shall give a brief outline of the theory 

 of capillarity in so far as it concerns the question at issue ; for the 

 pores are in effect merely fine capillary tubes. 



OUTLINE OF THE THEORY OE CAPILLARITY^ 



The general principle to be borne in mind is that the rise of 

 liquid in any capillary tube is primarily a measure of the pressure 

 discontinuity at the curved free surface of the liquid within the 

 tube. 



It can easily be shown that the pressure difference (A^) between 

 the two sides of a curved surface separating two fluids is expressed 

 by the formula: 



A^ = .(l+l) (I) 



where a is the surface tension and p and p the radii of curvature 

 in two planes at right angles to one another. If the surface is 

 spherical, p = p and the expression takes the simple form 



A^ = — . (2) 



'Thus G. A. Hulett {Zeitschr. d. physik. Chem., XLII (1903), 359, who made 

 experiments of this type using a porous porcelain plate in which copper ferrocyanide 

 had been deposited, observed in one instance a height of no cm. of mercury; in this 

 case therefore there was a negative pressure, amounting to about half an atmos- 

 phere, acting upon the water close to the under side of the porcelain disk. Similar 

 experiments had previously been made by E. Askenasy, Verhand. naturh. med. Verein 

 Heidelberg, March, 1895. This process, it may be remarked, is the commonly accepted 

 mode of accounting for the rise of sap in trees. 



^ Capillarity is treated in any textbook of physics, though not always well. The 

 discussion of it by Tait in his Properties of Matter is exceptionally clear. 



