228 SCIENCE PROGRESS 



but when it reaches the top of the meniscus again appears to be 

 straight. Before using the apparatus, it was carefully calibrated 

 with water in the following way : The apparatus was first 

 thoroughly cleaned, and was then filled with water so that it 

 reached a position below the reference mark. The water was then 

 run up and down the capillary several times, and the apparatus 

 was set vertical with the plumb-line and the height read ; the 

 tube was again wetted, turned through i8o degrees, and the 

 height again read (always allowing for the liquid to drain and 

 reading with a faUing thread) ; next the tube was inverted and 

 the two readings were repeated. More water was added and 

 the readings were taken, and so on. To show the kind of values 

 obtained, the following figures represent the mean of five read- 

 ings in each position: 1-4378, 1*4422, 1-4384, 1-4391 cm. The 

 mean of these is i -4394, with a probable error of 0-03 per cent. 

 By proceeding in this way, it was found, for example, with 

 one tube, that the most constant results were obtained within 

 the region 0-4 to 1-3 mm. of the reference mark, and this part of 

 the capillary was therefore used in the actual determinations. 

 The surface tension of benzene was obtained at 20° C. in both 

 tubes, with the following results : tube i, T == 28-94 ; tube 2, 

 T == 28-88 ; and the value for the surface tension of water at 

 the same temperature was found to be 72-62. The results 

 generally are higher than those obtained by other observers 

 using the capillary tube method, as might be expected owing to 

 the use of the wider tubes. 



Tate made experiments upon the weight of a drop of water 

 falling from the end of a glass tube. He used tubes from o-i 

 to 0-7 inch in diameter, and made of thin glass ground to a 

 sharp edge. He came to the conclusion that the weight of the 

 drop is proportional to (i) the diameter of the tube ; (2) the 

 weight of the liquid which would be raised up in the tube owing 

 to the capillary ; and he found that the weight decreased with 

 rise of temperature. His first two conclusions may be stated 

 in the form mg = k.rT. 



Many textbooks and research workers describe methods 

 using capillary tubes without any references to the sharp edges, 

 and using the equation mg = 2'7rrT, i.e. making Tate's constant 

 k equal to 27r. Poynting and Thomson {Properties of Matter) 

 point out that the forces acting on the drop include the excess 

 pressure over the external pressure which is developed in the 

 drop owing to surface tension ; if the drop is cylindrical at the 

 top, the value of this excess pressure is T/r, and we then have 

 2'7rrT = mg + irrT — i.e., mg = irrTy or exactly half the previous 

 value. Rayleigh finds {Phil. Mag., vol. xlviii, p. 321) that the 

 equation mg = ySrT is sufficiently accurate for most purposes, 

 while J. L. R. Morgan (see below) finds mg = yg^rT. There is 



