360 Professor Sir James Dewar [Jan. 18, 



For the purpose of extending the study to larger films, the 

 cylindrical globes that were used last year for the examination of 

 40 cm. bubbles were employed. Plane films 56j cm., or nearly 

 two feet in diameter, were thereby obtained which lasted for 40 days, 

 and even more, after becoming entirely black. Fig. 1 shows a globe 

 fitted for this purpose. About half a litre of soap solution having 

 been run in, and the end of the tube A dipped into it, a bubble, 

 blown by pure air, was started and gradually expanded, as shown by 

 the various stages of the diagram, until it became a plane horizontal 

 film about the middle of the vessel. Some idea of the appearance of 

 the film in the globe can be gathered from Fig. 2, reproduced from 

 a photograph taken with a white screen placed behind and a little 

 above the film, thus reflecting diffused light. 



The globe and its fittings must be scrupulously clean, and the 

 interior free from all traces of floating matter when tested by the 

 Tyndall cone of light. By careful manipulation the soap solution 

 was made to flow all over the inside, while the tube A was also wetted 

 by a stream of soap solution from a dropping funnel F fitting tightly 

 in the rubber cork used to close the neck of the globe. The tube A 

 can slide freely up and down in the support tube B, to which was 

 sealed the bulbed outlet E. The upper part of A could, on occasion, 

 be fitted with a dropping funnel C, sealed to the vertical portion, 

 while still connected to the bulbed nozzle D, by which pure air was 

 admitted. The bulbs D, E were loosely packed with cotton wool 

 moistened with glycerin. C was only used when subsequent additions 

 had to be made to the soap solution in the globe. 



After the plane film was obtained in the right position midway 

 in the globe, A was raised until its lower end was a few inches below 

 the film. The nozzles 1) and E were open, but were provided with 

 soda-lime guard tubes, so that the spaces above and below the film 

 were equally open to the atmosphere, though guarded from con- 

 tamination. This ensured that the fiuctuations of the atmosphere 

 were equalised throughout both portions of the relatively large space 

 of the globe ; otherwise the film was not likely to last long because 

 of the vertical displacements that would occur. In one case a baro- 

 metric alteration of 1 cm. in 20 hours caused the film to move 

 through r> mm. After the globe has drained for a day or two, there 

 is not enough liquid left on the walls to keep the film sufficiently 

 iuljricated to allow of such oscillations. The only liquid available is 

 in the tiny channel — the Gibbs layer — round the periphery of the 

 film. This channel is roughly triangular in section, with the base of 

 the triangle on the glass walls, and the concave sides converging to 

 the film which extends out beyond the apex of the triangle. In a 

 very well-drained film the base of this triangle may not be more 

 than O'l mm., so that the sectional area of the channel is about 

 3^^ Jooth of a sq. cm. The total liquid available round such a film 

 5G cm. diameter is, therefore, only about 1 cgm. It is obvious that 



