upon Surface-Films* 367 



In no case could any difference be detected in the behaviour 

 of camphor fragments on the two sides. Whenever possible, 

 the quantities of oil were adjusted to the point at which the 

 movements were just ceasing. In case of overshooting the 

 mark, the excess of oil could be easily removed by strips of 

 paper, partially immersed and then withdrawn, the action 

 being equivalent to an expansion of the surface. In several 

 cases the volatility of the substance with which the surface 

 was contaminated led to a subsequent retraction of the line of 

 dust. Thus freshly distilled oil of turpentine, even at first 

 barely capable of arresting the movements, soon passes off. 



As was shown by Tomlinson, oil of anise is incapable of 

 arresting the camphor movements. In the experiment with 

 a partition of dust, olive-oil will drive oil of anise into a very 

 small space, whose area is doubtless dependent upon the 

 amount of other impurities present. In this case, as in all 

 others, the behaviour of camphor is the same on the various 

 parts of the surface. 



It may thus be taken as established that the relation of a 

 contaminated surface to the camphor movements is one of 

 surface-tension only. 



A similar method of experimenting may be applied to a 

 rough determination of the degree of purity of cleansed sur- 

 faces. The whole of the surface under test is lightly dusted 

 over, and olive-oil is applied at several places close to the 

 circumference until camphor movements are nearly arrested. 

 After each addition of oil the dusted area contracts, and at 

 the close of operations it gives a measure of the extent to 

 which the original contamination must be concentrated in 

 order to stop camphor. 



A few numbers may be given as examples, although in all 

 probability the result is influenced by a variety of circum- 

 stances. A circular area of 10 inches diameter, occupied by 

 tap water, and cleansed by the flexible hoop described in 

 former papers, was tested on July 28, 1891. The application 

 of oil, just sufficient to stop the camphor movements, drove 

 the dust into a central circular patch of 2| inches diameter. 

 When the surface was in its natural condition, unpurified by 

 the action of the hoop, the central patch was of about 5 inches 

 diameter. These numbers, approximately verified on repe- 

 tition, show that the natural surface was about 4 times, and 

 the purified about 16 times better than according to the 

 camphor standard. The difference between the two cases is 

 less than was expected, and would perhaps have been greater 

 had distilled water been employed. It must be remembered 

 also that contact with dust (sulphur) is unfavourable to the 



2C2 



