212 ANNUAL KEPOET SMITHSONIAN INSTITUTION, 1912. 



knowledge which Plateau has bestowed upon us, it seems to me that 1 

 have but picked up a few crumbs which fell from the rich man's table. 



The formation and existence of a soap bubble depend upon the 

 weak superficial tension of the solution of soap and upon the remark- 

 able property, studied by Willard Gibbs, by virtue of which the 

 superficial tension varies, according to the needs of the moment, 

 between elastic limits, in the true sense of the term. I was sur- 

 prised to find by experiment that an increase amounting nearly to 

 20 per cent of the normal tension could be produced. Unfortunately, 

 I can try no experiment here wliich would permit me to demonstrate 

 to you this fact adequately. 



The tension at the upper parts of a large bubble must be greater 

 than that at the lower parts, for it must balance both the weight 

 of the bubble and the tension of the lower parts. That is why there 

 must be a superior limit to the possible size of a soap bubble. A 

 bubble whose color is the bright apple green, weighing five one- 

 thousandths grain per square inch,^ can not exceed 100 inches in 

 diameter, for in that case the additional force which the upper part 

 has to withstand is one-half grain per linear inch, which is one-fifth 

 of the normal tension of the film, i. e., 2^ grains per linear inch. 

 Similarly a bubble white of the first order, or weighing one one- 

 thousandth grain per square inch, might extend to five times this 

 diameter before this cause of failure would operate. But it is not 

 possible in practice to blow such large bubbles. One great difiiculty 

 is, if mechanical means be not employed, to send air in sufficient 

 quantity. With regard to tliis. Prof. Wood, of Baltimore, told me 

 that he found the principle of injection very advantageous. Indeed, 

 when we reflect, it exactly meets the necessities of the case. Internal 

 pressure diminishing as the bubble increases in size, a small quantity 

 of air blown into the tube will carry ^vith it a large amount of air 

 at a small but sufficient pressure. 



I have tried several forms of injectors, but the simplest and hitherto 

 the best is made of a bent pipe such as is employed in the testing 

 of illuminating gas for sulphur. I blow into the narrov/ end by 

 means of a mouthpiece, while the wide end is surrounded by a cambric 

 band with a serrated edge which feeds the bubble with liquid as -it 

 increases. With tliis device I have not only blown a bubble of 80 

 centimeters diameter, but am convinced that this is by no means 

 the practical limit. It may be worth mentioning that I make use 

 of Plateau's liquid consisting of a solution in water of oleate of soda 

 with glycerine added. The proportions are as follows: Oleate, 1 

 part; water, 40 parts; one-third of its volume of glycerme is then 

 added. I have increased the proportion of oleate, especially when 

 I have wished to blow large bubbles. 



« These weights can be read ofl directly from the colored plate in my book. 



