432 SCIENCE PROGRESS 



therein of a liquid of known surface-tension. But this method 

 of procedure at once raises the question of the validity of 

 whatever method has been used for the determination of the 

 surface-tension of the standard liquid. 



Further, the thorough cleaning and drying of very narrow 

 capillary tubes are matters of no small difficulty, whilst, if the 

 measurements be carried out over any very wide range of 

 temperature, the steadily decreasing value of the rise — which 

 becomes zero at the critical temperature — subjects the measured 

 value of h to a constantly increasing percentage error. 



These latter defects are quite avoided in Wilhelmy's method 

 (No. 15 in Table I), in which surface-tensions are determined 

 by hanging a number of plates beneath the pan of a balance 

 with their planes vertical and their lower edges in a horizontal 

 plane. The liquid under examination is then adjusted until its 

 surface just touches the lower edges of the plates, when the 

 downward pull on the plates due to surface-tension may be 

 estimated by adding weights {mg) to the other pan of the 

 balance. If p be the perimeter of the plates, the surface-tension 

 of the liquid is given by mg = pT, if we assume the contact- 

 angle to be zero. The general equation, assuming a finite 

 contact-angle, is mg = pT cos 6. 



This method again, in common with the capillary-rise 

 method, suffers from the fatal defect that a knowledge of the 

 contact-angle 6 is necessary before T can be determined. But, 

 apart from this, it seems strange that the method should not 

 have been more generally used in physico-chemical researches. 

 For, depending as it does on a weight-determination, it can be 

 made far more sensitive than the capillary-rise method, which 

 depends on the exact estimation of a length, whilst, for liquids 

 of very small surface-tension, its sensitiveness can be increased 

 to any extent desired by increasing the perimeter of the plates 

 touching the liquid surface. Further, the plates can be cleaned 

 much more easily than can a glass tube of capillary bore. 



The other methods classified in section B need not be re- 

 viewed in detail. Whilst, in the laboratory, they possess great 

 educational value both as providing experimental illustrations of 

 points in capillary theory, and as giving useful exercises for 

 laboratory practice, considered as instruments of research they 

 all possess the inherent defect that they demand a knowledge of 

 the contact-angle before they can be used to give reliable values 



