﻿of the common Surface of two Liquids, 473 



and the magnitude of the drops again increases. Owing to the 

 difficulty of estimating the thickness of the coating of oil cor- 

 rectly, I have made no measurements with regard to it. 



The longer the drops are in falling, the more time have they 

 to condense substances on the surface which were contained in 

 the atmosphere in the form of vapour, and just so much less must 

 a or the capillary constant be found to be, as experiment also 

 teaches. In the earlier* methods which I used with falling 

 drops for determining the capillary constant of fused substances, 

 this source of error had, on account of the high temperature of 

 the fused substances, only a slight influence. The same holds 

 good for measurements of flat drops of fused substancesf. 



If a horizontal capillary tube of radius r be fastened perpen- 

 dicularly to the axis of an ordinary reflecting goniometer, and 

 a small quantity of a liquid 1 (say, water) be placed in it, then 

 the thread-like column of liquid between the two capillary me- 

 nisci will be in equilibrium in all places of the cylindrical tube. 

 If now a small quantity of a liquid 2 be placed upon one me- 

 niscus of the column of liquid 1, then by these means the capil- 

 lary pressure — - on the top of the same will be diminished to 

 iq — ; and the column of liquid pushed towards the other 



end of the capillary tube. By a certain inclination v, measured 

 on the goniometer, of the capillary tube, which depends on the 

 length L and the specific weight a i of the column of liquid 1, 

 the column of liquid will again stand in equilibrium. Such a 

 column of liquid with heterogeneous surfaces in a cylindrical 

 glass tube behaves similarly to a column of liquid with homo- 

 geneous surfaces in a conical glass tube. The same difficulties 

 as in the experiments described in Sections 3 and 4 stand in 

 the way of obtaining a valuefora 12 + a 2 by measurements of the 

 magnitudes v, L, and r, in that the changes of the angle cannot 

 be estimated with sufficient facility. For lecture-room experi- 

 ments, on the contrary, this method is well adapted to show 

 the dependence of the capillary pressure or the tension of sur- 

 face on the nature of the free surface of the liquid. 



If a liquid 1 (say, water) be allowed to rise in a capillary tube 

 in the usual manner, and then small quantities of a liquid 2 (say, 

 olive oil or oil of turpentine) spread on the free plane surface of 

 the liquid outside the capillary tube, then the elevation of the 

 liquid in the capillary tube remains unchanged. 



As the constant oc l or Hj (compare § 1) is Very essentially 

 modified by the introduction of this liquid 2, I should also have 



* Pogg. Ann. vol. cxxxv. p. 642 (1868). 

 t I J ogg\ Ann. vol. cxxxviii. p. 150(1869). 



