434 SCIENCE PROGRESS 



And although this difficulty is not insurmountable, especially 

 in the case of the ripple-method, some of the methods afterwards 

 to be discussed are more accurately statical and permit of 

 greater compactness of arrangement of the liquid under ex- 

 amination, and hence allow of more accurate observation and 

 control of the temperature. 



Turning now to the weighing methods, (4) determines 

 surface-tensions in terms of the mechanical pull on a ring 

 formed from a piece of thin rod bent into a circle of large radius. 

 If the anchor ring thus formed be suspended underneath the 

 pan of a balance with its plane horizontal, be allowed to touch 

 the surface of the liquid under observation, and then be slowly 

 withdrawn, a distinct maximum pull is observable as the ring 

 is raised from the liquid surface. From a knowledge of the 

 dimensions of the ring and of this maximum pull the surface- 

 tension of the liquid can be calculated with considerable 

 accuracy, if the dimensions of the ring be properly chosen. 

 It will readily be seen that the principles employed in the 

 approximate integration of the differential equation to the 

 capillary surface which give the working formulae for this 

 problem are of much the same type as those which govern 

 the formulae obtained in the case — to be discussed immediately 

 — of large bubbles and drops; and when it is said that the 

 anchor ring must be of large radius, the word " large " is used 

 in exactly the same sense in both cases. As far as exactness 

 is concerned, the method leaves little to be desired, and the 

 only factor that has prevented it from coming into wider use 

 is, perhaps, the difficulty of hitting off the maximum position 

 exactly, and some lack of appreciation of the exact conditions 

 under which the approximate formulae used in the computations 

 may be applied. 



In the practice of method (5) a light frame is made by 

 bending a piece of thin glass rod so as to form three sides of 

 a rectangle. The frame is suspended from the arm of a balance, 

 dipped into the liquid under examination, and raised therefrom 

 until a film is formed which extends from the horizontal bar to 

 the liquid surface. The balancing weight is now read off, and 

 again after the film is broken, when the difference of the two 

 weights divided by twice the width of the film gives the required 

 surface-tension. Subject to one or two small corrections, which 

 are easily made, this method is both sensitive and convenient, 



