1917] on Soap Bubbles of Long Duration 189 



purpose of starting the bubble. The reservoir is shut off bv a stop- 

 cock from the blowiuii' tul)e, as it is not advantageous for the soap 

 sokition to be exposed to the current of air used in expanding the 

 bubble : a side tube is sealed in for the air inlet. The blowing tube 

 is mounted, so that it can slide up and down while remaining air- 

 tight. A few small bubbles are first blown and run round the 

 supporting ring to wet it. A bubble is then blown to about 4 cm. 

 greater diameter than the ring, while resting on one side of it. This 

 is done bv holding the end of the blowing tube sufficiently out of 

 the centre, and steadily raising it until the bubble is about 15 cm. 

 diameter, when it is allowed to come into full contact with the whole 

 ring, and the expansion is continued until the requisite diameter of 

 about 40 cm. is reached. Meanwhile, the blowing tube is continually 

 raised through its sliding support tube in the india-rubber cork, and 

 finally Avithdrawn by a to-aiid-fro spiral movement upwards. In one 

 case when the bubble became free from the tube, a complex up-and- 

 down oscillation took place in a period of about two seconds, and with 

 an initial amplitude approaching 3 cm. at the upper periphery. By 

 leaving the blowing tube in the bubble the advantage of an added 

 steady support is secured ; this is unsatisfactory when measures of 

 contraction are being made, but it certainly helps to prevent undue 

 disturbance, especially in the early stages, when the bubble is liable 

 to be thick and somewhat top-heavy on the 10 cm. ring. 



It can be seen from Table 2 that there is a general increase in 

 length of life of the successive bubbles in both the first and second 

 vessels. The last items, in both cases, snow a shorter life ; this, 

 however, was due in one case to the vessel being in the open air in 

 winter, whereby its temperature at the end went below freezing 

 point : and in the second case, to the globe becoming contaminated, 

 from the displacement of a badly-fitting tube. The irregularity 

 between the fourth and fifth items was explained by the hot weather 

 (temperature over 20° C). 



The maintenance of colour in such large bubbles was not easy. 

 For this purpose a good circulation of water vapour in the vessel seems 

 desirable, as the mere presence of a litre or two of water in it was 

 not entirely effective. A wet cloth placed on the upper part of the 

 vessel (seen in Fig. 4) by its own evaporation maintained a slightly 

 cooler area, thus promoting the condensation there of drops of water. 

 A flow of water vapour up through the vessel was thus obtained, 

 which being partly absorbed by the glycerine in the bubble did for 

 a time maintain, and even increase, its thickness and colour. In the 

 later stages also, when the bubble material has thus become altered in 

 the relative proportions of its constituents, vaporisation from the 

 bubble may take place, and then the opposite effect is observed ; 

 distillation causing the thickness to deciea-e, so that two or three 

 colours are passed through in as many hours. A few degrees rise of 

 room temperature causes the same change, especially in diluted 



